This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kwaku, K. F. Articles by Dillon, S. M. Search for Related Content PubMed PubMed Citation Articles by Kwaku, K. F. Articles by Dillon, S. M.

(Circulation Research. 1996;79:957-973.)
© 1996 American Heart Association, Inc.

Articles

Shock-Induced Depolarization of Refractory Myocardium Prevents Wave-Front Propagation in Defibrillation

Kevin F. Kwaku, Stephen M. Dillon

the Department of Pharmacology (K.F.K.), College of Physicians and Surgeons, Columbia University, New York, NY, and Philadelphia Heart Institute (S.M.D.), Sidney Kimmel Cardiovascular Research Center, Presbyterian Medical Center, Philadelphia, Pa.

Correspondence to Stephen M. Dillon, PhD, Allegheny University of the Health Sciences, Division of Cardiology, mail stop 429, Broad and Vine, Philadelphia, PA 19102-1192. E-mail sdillon@ix.netcom.com.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The elimination of most, if not all, propagating wave fronts of electrical activation by a shock constitutes a minimum prerequisite for successful defibrillation. However, the factors responsible for the prevention of postshock propagating activity are unknown. We investigated the determinants of this effect of defibrillation shocks in 23 Langendorff-perfused rabbit hearts by optically mapping cardiac cellular electrical activity by means of laser scanning. The optical action potentials obtained by this method were continuously recorded from 100 ventricular epicardial sites before, during, and after shock delivery during fibrillation. Analysis of activation maps showed that postshock propagating activity arose from areas depolarized by the shock. In 273 shock episodes, 898 sites at the border of shock-depolarized areas (BSDAs) from which wave-front propagation could have arisen were identified. The incidence of postshock propagation from BSDA sites was inversely related to refractoriness, as indexed by coupling interval (CI) or the optical takeoff potential (V_m). Specifically, there was a near-zero probability of postshock propagation if the shock caused depolarization at CIs <50% of the fibrillation cycle length or from myocardium still depolarized to 60% of the amplitude of a paced action potential (APA). Furthermore, incidences of wave-front propagation following shocks were consistently lower than the propagation incidences of naturally occurring unshocked fibrillation wave fronts, at comparable CIs and V_ms. We conclude that the incidence of postshock wave-front propagation decreases with increasing refractoriness at the BSDA and that shock-induced depolarization of effectively refractory myocardium (ie, depolarized to 60% APA) is required to guarantee the cessation of continued wave-front propagation in defibrillation.

Key Words: defibrillation • ventricular fibrillation • refractoriness • optical mapping • rabbit heart

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ventricular fibrillation is a complex and disorganized self-sustaining reentrant cardiac rhythm^{1 2} responsible for hundreds of thousands of fatalities a year in the United States alone. Delivery of an electric shock to the heart is the only effective therapy for VF. Although the technique has been the object of experimental study for nearly 100 years³ and has been applied clinically for over 50 years,⁴ there is still no adequate description of how defibrillation is achieved. In recent years, most of the new empirical findings concerning defibrillation have been considered in the context of two different theories of defibrillation, the CM^{5 6} and ULV^{7 8} hypotheses (see References 9 through 12 for reviews). The CM hypothesis holds that defibrillation is accomplished through the annihilation of propagating activity in a CM of the heart, which can be <100%. By contrast, the ULV hypothesis posits that shocks that stop all ongoing VF wave fronts can fail to defibrillate, because these shocks also induce one or more new reentrant circuits, which immediately refibrillate the heart. Induction of this reentrant activity is thought to occur via a process known as the critical point mechanism. Despite significant differences between the CM and ULV hypotheses, both consider the prevention of wave-front propagation by a shock to be central to defibrillation. The CM hypothesis^{5 6} did not postulate a particular mechanism for the elimination of postshock propagating activity by a shock. However, some investigators^{11 13 14} have attributed blockade of postshock wave fronts in processes underlying the ULV hypothesis to the production of the so-called graded response.¹⁵

The purpose of the present study was not to test either the CM or ULV hypotheses specifically but rather to address the more general problem of how defibrillation shocks stop wave-front propagation. We are aware of no study that has expressly and quantitatively investigated the mechanistic determinants of postshock propagation in the setting of defibrillation. We have used optical mapping by laser scanning^{16 17 18} to record transmembrane potential–dependent optical signals at 100 epicardial sites in isolated perfused rabbit hearts. Recordings were made before, during, and after defibrillation shocks. In contrast to electrical mapping, the optical technique is immune to shock-induced signal artifact and interruptions, senses nonpropagating as well as propagating electrical activity, and portrays the time course of membrane potential changes. This approach allowed us to continuously assess the effect of shocks on patterns of fibrillation wave-front propagation as well as to make quantitative measurements of the underlying electrophysiological responses to shocks at each site. We were able to identify sites from which postshock propagation might arise and found that the likelihood of such activity occurring was strongly dependent on the level of refractoriness prevailing at these sites at the time of the shock. Postshock propagating activity was abolished when the shock was able to cause depolarization in myocardium already in its effective refractory period. Some of these results have been presented in preliminary form.^{19 20}

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Experimental Preparation
Experiments were carried out on a total of 23 isolated intact hearts obtained from New Zealand White rabbits of either sex weighing 3.17±0.29 kg (mean±SD; range, 2.64 to 3.86). Animal care and experimental procedures were approved by the Institutional Animal Care and Use Committees of Columbia University and Presbyterian Medical Center. Rabbits were intravenously injected with 1000 to 5000 U heparin and 40 to 50 mg/kg body wt pentobarbital. The hearts were rapidly removed and immediately immersed in cold Tyrode's wash solution. The hearts were then mounted onto a Langendorff apparatus, and the coronary system was perfused with warm oxygenated Tyrode's solution via a cannula secured in the aortic root. The left ventricle was flushed of blood with Tyrode's solution injected through the mitral valve. The Tyrode's solution consisted of the following (mmol/L): NaCl 130, NaHCO₃ 24.2, KCl 4, CaCl₂ 1.8, MgCl₂ 0.6, and NaH₂PO₄ 1.2. Included in the perfusion solution, but omitted from the cold Tyrode's wash solution, were the following: 11.1 mmol/L dextrose, 2 mmol/L sodium pyruvate, 40 mg/L bovine albumin (Sigma Chemical Co), and 10 U/L pork insulin. The Ca²⁺ channel blocker D600 (methoxyverapamil, Sigma) was added in most experiments (final concentration of 2 µmol/L) to suppress cardiac contraction. The perfusate was continually gassed with a 95% O₂/5% CO₂ mixture to give a pH of 7.35 to 7.40. The solution was prewarmed to 30°C to 32°C in a glass beaker placed on a heating plate located 40 cm above the heart. Tyrode's solution was brought to the heart via stainless steel tubing to prevent diffusional gas loss and warmed again by a heating coil just before entering the heart. Solution temperature was regulated by a thermostat so as to maintain a left ventricular cavitary temperature of 36±1°C, measured by thermocouple probe placed inside the ventricle. A warming jacket was not used in order to maintain optical access to the heart. Consequently, the right ventricular epicardium was 1.2°C cooler than the left ventricular cavity, resulting in epicardial conduction velocities only marginally slower (4%) than if no temperature gradient had existed.²¹ Conduction velocity was measured by the method of Schalij et al²² using a cross-shaped array of 16 extracellular unipolar electrodes (four electrodes per arm, 1 mm separation) and a single center stimulation electrode. The coronary effluent was collected and returned to the reservoir of Tyrode's solution; a total volume of 1.4 L was thus continuously recirculated. Except during fibrillation, the heart was paced at a basic cycle length of 300 ms through a single bipolar or, in nearly half the experiments, a linear array of four platinum unipolar electrodes placed on the ventricular epicardium. Unipolar stimulation current was returned through the cathodal defibrillation electrode. Pacing stimuli were hardware-fixed constant-current pulses of 4-mA amplitude and 2-ms duration, or 7.5 times the diastolic pacing threshold. The cardiac rhythm was continuously monitored using a cavitary ECG recorded between an electrode on the aortic root and one lying in the left ventricular cavity and displayed on a digital oscilloscope (model 1425, Gould Inc). All but two hearts were blotted and weighed at the conclusion of each experiment. Hearts used in the present study weighed 9.21±1.60 g (range, 7.0 to 13.16 g).

Optical Mapping Technique
Optical cardiac mapping was accomplished by means of a laser scanning technique.^{16 17} A detailed description of the technique is given elsewhere,¹⁸ but important technical specifications are included here. The fluorescent voltage-sensitive dye WW781 (dye XXV of Gupta et al,²³ available as W-435 from Molecular Probes, Inc) was added to the recirculating Tyrode's solution to obtain a final concentration of 4 mg/L (5.3 µmol/L). This dye binds to cell membranes and linearly transduces transmembrane voltage into a fluorescence signal. The dye responds only to transmembrane potential and remains insensitive to extracellular and intracellular voltage gradients generated by the shock, which can transiently saturate the amplifiers of electrical recording systems.^{24 25 26} We have previously used WW781 in this preparation to optically record membrane activity from single sites during defibrillation-strength shocks.^{27 28}

The layout of the laser scanning system is depicted schematically in Fig 1A. A beam of red light (647-nm wavelength, 400-mW power) emitted by a krypton ion laser (Innova 90K, Coherent) was used to elicit fluorescence from the voltage-sensitive dye. Because of optical losses within the scanning system, only 55 mW of laser light reached the heart. The laser spot (200-µm diameter) was moved from one of the 100 sites to another every 10 µs by a pair of acousto-optical deflectors (EFL D250, InRad) under computer control. The laser power reaching the heart in the present study was 5.5 times that of our previous optical studies^{27 28} using a fiberoptic system (continuous power of 10 mW into a 250-µm-diameter spot), but since the laser illuminated a given spot only 1% of the time, the effective dose of laser illumination at each spot during scanning was <1/10 of that during fiberoptic recording (power density, 1.75 and 20.4 W/cm², respectively). Therefore, there was no concern over the use of a more intense laser beam. In order to simplify the optics, we underfilled the deflector apertures and so traded off resolution from 250 discrete spots per axis down to 137 spots per axis. A total of 5.5 µs was needed to redirect the laser spot upon receipt of a new coordinate control signal because of the time required for the deflector drivers (DE-70M and DE-70BM, IntraAction) to settle to a new output and for the time needed by the new acoustic wave fronts to propagate to and through the laser beam in the acousto-optic deflectors. Laser scan coordinate control signals for each axis were generated by a pair of programmable waveform synthesizers (WSB-10, Qua Tech). The laser scanning system could repeatedly redirect the laser to a given location to within 5% of the minimum possible separation between laser spots (peak-to-peak error in position measured during a 1-s scan). In these experiments, we recorded from 100 sites on the ventricular epicardial surface and thus realized a sampling rate of 1000 samples per second per site. The fluorescence elicited from each site was passed through a long-pass filter (RG665, 3-mm thickness, Schott Glass Technologies) to reject reflected laser light and detected by a single photodiode (600-PIN-RM, Quantrad Sensor), typically placed 4 to 7 cm from the heart. The resultant photocurrent was amplified and converted into voltage. The photodetector system had a total gain of 5x10⁶ V/A and a 20-mV peak-to-peak dark noise level and could completely settle to a new fluorescence level in <10 µs after the receipt of a new laser coordinate control signal. Approximately 10% of the light leaving the laser was directed to a reference photodetector (see Fig 1). The signal from this detector was used to compensate the fluorescence signal for transient fluctuations in the laser power by continuous high-bandwidth analog division (MPY634, Burr Brown).

View larger version (120K): [in this window] [in a new window]   Figure 1. A, Schematic diagram of optical mapping setup. Light (647 nm) is emitted from a krypton laser and immediately passed through a red-pass filter (F1) to reject plasma emissions. Ten percent of the light is reflected by a beam-splitter (M1) to a reference detector used to compensate for laser power fluctuations. The remaining light is directed by a front-surface mirror (M2) to two acousto-optical deflectors (AOD1 and AOD2), which change the beam angle in the vertical and horizontal planes, respectively. A pair of lenses (L1 and L2) magnify and focus the deflected laser beam. The scanned laser beam (two positions are shown) is then reflected from a dichroic mirror (M3) and excites the fluorescent dye at a recording site on the heart. The longer-wavelength fluorescence returns toward M3, is transmitted by the mirror, and is collected by a single photodiode-type fluorescence detector for transduction into a voltage signal. B, Reflectance image of a heart with numbered recording sites superimposed. C, Drawing of the heart showing location of shock electrodes, ECG recording electrodes, and temperature probe.

The system could scan a maximum area of 24.7 mmx24.7 mm with this optical configuration, although the effective scan area was generally limited to 20 mmx20 mm because of the hearts' curvature at the far edges of the scan area. The typical interrecording site distance was thus 2 mm. The area to be scanned was delimited using custom data acquisition software and automatically filled with 100 sites laid down in a regular pattern so that each site within the scan area had four or six equidistant nearest neighbors. The position of some sites was often manually edited to move sites yielding poor recordings due to the curvature of the heart or epicardial fat. For example, Fig 1B shows a reflectance image¹⁸ of a heart (upon which is superimposed the locations of recording sites numbered 0 through 99) obtained using the scanning system before the addition of dye and without the long-pass filter. Recording sites were positioned within the area delimited by the experimenter. Sites 91 through 99 interrupt an otherwise regular pattern of recording sites because they have been manually repositioned. The pattern of scan sites was unique to each experiment. Recordings were made from the right (n=21) and left (n=3) ventricles or both at once, in which case recording sites were centered on the posterior descending artery (n=2) (some hearts were scanned on more than one surface).

Data were acquired and digitized with 12-bit resolution (DT2821-G-16SE, Data Translation) and displayed by the custom software running on an IBM-compatible 486DX personal computer. This is the same software that controlled the positioning and movement of the laser spot. A timing signal was added to the analog-to-digital data stream to register the timing of events such as shocks and pacing stimuli. The optical recordings in the present study have had the baseline fluorescence signal component subtracted and have been filtered using a 5-ms window sliding average algorithm.²⁷

Fibrillation and Defibrillation
The crosshatched regions of Fig 1C depict the location of basal band and apical cup electrodes used for electrical fibrillation and defibrillation.²⁷ These electrodes were made of stainless steel mesh (Small Parts Inc). The apical cup served as a cathode for all shocks and had a surface area of 398 mm²; the basal band anode was 5 mm or 7 mm wide, and its length varied with ventricle circumference to yield an effective surface area of 350 to 700 mm². VF was electrically induced by passing a 60-Hz alternating current between these electrodes. Test and rescue shocks were generated by two battery-powered defibrillators (models 2394 and 2326, Medtronic Inc). This study used monophasic truncated-exponential waveforms of 5-ms duration and a 58% average tilt for the test shock and 4- to 6-ms duration and 63% tilt for the rescue shock. Test shocks were delivered at various strengths. A resistance of 212 was placed in the discharge circuit between the defibrillator and heart, so that only 27±9% of the voltage dialed on the defibrillator, depending on cardiac impedance, reached the heart. (A resistance was added to the shock path in order to increase the resolution of the shock voltages selected on the 2394 defibrillator and because the lowest shock strength the test defibrillator could reliably generate was a dialed voltage of 60 V.) The average cardiac impedance was calculated on the basis of Ohm's law to be 78 . Rescue shocks were delivered directly to the heart, bypassing the resistance in the test shock circuit, and were set to a strength sufficient to consistently defibrillate: 174 to 266 V.

The diagnosis of the induced arrhythmia as VF was routinely based on the cavitary ECG. Fig 2 shows the ECG and an optical trace of membrane voltage (panel A) and isochronal maps of activation patterns (panels 2B through 2E) from a typical VF episode. The ECG complexes are rapid, irregular, and polymorphic, albeit more organized in appearance than VF typically recorded from the body surface, perhaps because of the proximity of the ventricular myocardium to the cavitary electrode. The optical trace below shows action potentials with characteristics similar to those reported by microelectrode recording during VF in rabbits,²⁹ dogs,^{30 31} and swine,³² as well by optical recording in rabbits²⁸ and swine.³³ In particular, the action potentials vary considerably in amplitude, duration, takeoff potential, and overall morphology. Moreover, the action potentials are not phase-locked with the ECG complexes, as might be expected during a polymorphic but repetitive rhythm. The accompanying activation maps yield the most compelling evidence for the phenomenon of VF in this example. Panels B through E of Fig 2 show isochronal maps of right ventricular epicardial activation patterns during the time intervals indicated by the lettered bars above the ECG in panel A. Only the earliest and latest isochrones (drawn at 10-ms intervals in these and subsequent maps) of each wave front are labeled for simplicity, and shaded arrows trace the paths of each numbered wave front. Over the course of 1 s, very different activation patterns, which depict many characteristics seen in other mapping studies of fibrillation in both the atria³⁴ and the ventricles^{25 35 36 37 38} of large animal hearts, can be seen. The majority of the time there were multiple wave fronts present within the scan area, and these wave fronts followed complicated paths imposed on them by transient and shifting lines of block (compare the positions of the lines of block in each map). Wave fronts (referred to by panel letter and numerical designation) were characterized by splitting into daughter wavelets, (eg, B4, C1, C2, D3, and E2), collision and mutual annihilation (eg, B2 and B3; E1 and E2 [upper parts]; E3, E4, and E5), and fusion of two wave fronts into one (eg, C2 and C3, E1 and E2 [lower parts]). Panel C of Fig 2 shows evidence of transient organized reentry (two revolutions) in the upper portion of the maps (C2 represents the continuation of C1). Such circuits never persisted for more than a few revolutions, and multiple, simultaneous, complete reentry loops were much more rare.

View larger version (113K): [in this window] [in a new window]   Figure 2. Temporal and spatial activation dynamics of VF in the rabbit heart. A, Cavitary ECG obtained from the left ventricular apex showing polymorphic morphology (top trace) and simultaneously recorded fluorescent signal from a single right ventricular epicardial site (bottom trace). Note the beat-to-beat changes in APA, action potential duration, Vm, and shape. B through E, Isochronal maps corresponding to the times indicated by the lettered bars at the top of panel A. Recording sites are indicated by small dots; for simplicity, individual activation times (as used in subsequent figures) have been omitted. The picture is dominated by multiple wave fronts conducting around shifting lines of block, resulting in complex disorganized activation patterns characterized by wave front splitting, merging, and reentry. However, during some intervals, simpler dynamics exist. See text for details.

Despite the generally complex dynamics overall, there were time intervals in this episode when the dynamics were simpler, eg, when there was only one wave front (eg, 120- to 170-ms interval in panel B of Fig 2) or when wave fronts were relatively planar. There was also a 40-ms gap between the last activation in the top map and the earliest activation in the bottom map of panel E during which no wave fronts were detected at all; although this was not the norm, such a pause was not unusual. In the examples presented later in the present study, we have selected episodes in which the activation patterns immediately following a shock were roughly planar in order to most clearly illustrate the effects of shocks on wave-front propagation.

Experimental Protocol
The experimental protocol is depicted schematically by the time bar in Fig 3. VF was induced by 5 s of 60-Hz alternating current stimulation. Fibrillation was allowed to proceed for 10 s, as documented by the cavitary ECG, before a 1- to 1.3-s laser scan was initiated. Coronary perfusion continued throughout the period of fibrillation. Our earlier studies indicated no qualitative influence of a short period of ischemia on the electrophysiological effects of a shock.²⁸ Stopping perfusion caused a slow change in heart shape as vascular pressure subsided, an effect that would prevent accurate registration of the recording sites.

View larger version (35K): [in this window] [in a new window]   Figure 3. A, Time bar depicting sequence and duration of experimental protocol (top) and representative optical voltage trace from one of 100 sites (bottom). Test shock, rescue shock, and pacing stimulus were delivered at times indicated by the arrowheads. In this example, CI was 46 ms, or 70.4% of the mean VFCL, which was 63.3±6.5 ms. Vm was 35.6% of the control APA. B, Example, from the same heart as in panel A, of a shift in baseline fluorescence signal associated with the rescue shock. Although infrequent and found at the edges of the heart, including the apex, such shifts precluded the accurate determination of the Vm at these sites.

The trace below the time bar shows a typical recording of the optical membrane voltage from one of 100 recording sites. The optically recorded fibrillation action potentials show characteristics similar to those reported using transmembrane microelectrode recording.^{29 30 31 32} In particular, VF action potentials varied considerably in both amplitude and duration, and action potential upstroke takeoff potentials were depolarized above the resting potential. Uninterrupted VF was recorded for a minimum of 300 ms, after which a test shock was delivered. In the example in Fig 3, the test shock (150 V dialed, or 40 V to the heart) elicited a premature depolarization (ie, occurring at a coupling interval 2 SD below the mean VF cycle length) but failed to defibrillate. Postshock activity was recorded for a minimum of 350 ms, after which a rescue defibrillation shock (174 V to the heart) was delivered. The rescue shock caused a large depolarization, fibrillation ceased, and the transmembrane potential smoothly repolarized to a steady resting level. A stimulus was delivered from the remote pacing electrodes 200 ms after the rescue shock, and a propagated control action potential was recorded. Note that neither the test nor the rescue shock interrupted the optical recording. In a small number of episodes, there was a shift in the baseline signal corresponding to resting membrane potential following a shock, which prevented measurement of the control APA. An example of this phenomenon is given in Fig 3B, taken from the same heart as in Fig 3A. Such shifts were most prominent at the apex and edges of the heart after rescue shocks and were therefore probably motion related; however, the mechanism by which they occurred is not known. It is possible that Ca²⁺ entered cells through pathways not blocked by 2 µmol/L D600, such as reverse-mode Na⁺-Ca²⁺ exchange.^{39 40}

In most experiments, test shock strengths were initially set to low voltages (90 to 150 V dialed on the defibrillator) and then incremented in steps of 30 V until defibrillation was observed in two succeeding episodes. In six experiments, shock strength was then decreased, again in 30-V decrements, back to beginning shock strengths. In one experiment, shock strengths were alternated between randomly selected high and low levels. Dialed test shock strengths were in the range of 90 to 510 V, which resulted in a range of 24 to 137 V applied to the heart. This protocol was chosen because it was discovered early in our studies that successful defibrillation was often associated with a shock-induced depolarization throughout the entire scan region that prevented immediate postshock propagation, the latter being the object of our study. A minimum of 2 minutes was allowed to elapse between each shock episode, during which the heart was continually paced at a 300-ms basic cycle length. Fig 4 shows the probability of successful defibrillation as a function of test-shock dialed voltage (binned in 60-V increments) combined for all test shocks delivered to all hearts in the presence of D600. The data were fitted by a four-parameter logistic equation (SigmaStat, Jandel Scientific) constrained between 0% and 100% defibrillation success, which rendered the well-known sigmoid relationship between probability of success and shock strength.⁴¹ The absence of data points along the maximum asymptote reflects the design of our protocol, which did not include the application of test shocks much stronger than those observed to defibrillate.

View larger version (15K): [in this window] [in a new window]   Figure 4. Probability of defibrillation versus test shock voltage dialed on the defibrillator for 273 shocks in 23 hearts. The line represents best fit to the data by the following four-parameter logistic equation: y=(a-d)/[1+(x/c)b]+d. The fit shows that the data follow the expected sigmoid relationship (R2=.982). The lack of data points at the plateau reflects the fact that shocks much stronger than needed to defibrillate were not delivered. Note that an average of 27±9% of the dialed voltage was delivered to the heart.

Data Analysis
Each optical recording was analyzed to yield information about the timing of the test shock relative to the electrical state of the myocardium at the moment of shock delivery. The CI of the test shock at a specific site was computed as 100% times the interval between the shock-induced depolarization and the preceding VF activation, divided by the mean VFCL. VFCL was calculated for each fibrillation episode as the average of all interbeat intervals recorded at all sites during VF before the test shock (n=331±79 intervals per episode). An index of the membrane potential at the time of the shock was also measured. Since voltage-sensitive dyes yield measurements of change in transmembrane voltages but not the voltages themselves, we report the optically indicated potentials from which test shocks induced a depolarization as 100% times the takeoff level above resting potential, divided by the control APA, for normalization between sites. We refer to this value as V_m. APA was measured between the resting level and the peak of the paced action potential.

The primary method of analysis was the construction of activation maps depicting the propagation of depolarization before, during, and after the application of the test shocks during VF. Activation times in each laser scan were initially marked by a computer algorithm, followed by manual correction of inappropriate marks. Activation times were taken as the temporal midpoint of depolarizing deflections, whether these were due to ongoing fibrillation activity or due to shocks. Depolarizations were considered to be shock-induced if they met the following criteria: shock-induced depolarizations (1) began during the shock, (2) had the largest rise and fastest rate of rise within 15 ms of shock onset, and (3) peaked within 20 ms of shock onset. In rare cases in which the optical recording showed an initial hyperpolarization, the immediate postshock potential was required to be more depolarized than the preshock potential for shock depolarization to be scored, which, with few exceptions, was the case.

Isochronal activation maps were constructed from activation times at all sites, plotted with respect to the time of the test shock. The criteria for indicating impulse block of naturally occurring impulses were the same as those given below for the propagation block of shock-induced depolarizations. A typical example illustrating our methods of analysis is shown in Fig 5. Panels A and B respectively show maps of impulse propagation just before and after a test shock. Activation times were plotted with respect to the leading edge of the shock waveform; therefore, preshock maps show negative activation times. Panel A shows a single fibrillation wave front at the top of the scan region (base of the ventricles) 79 ms before the shock, which propagated toward the apex, where it exited the scan region 29 ms before the shock. The next VF wave front, in panel B, again entered at the top of the map. At time 0, a shock was applied, which resulted in a broad isochronal band of activation ahead of the preshock VF wave front (area between isochrones 0 and 10). Subsequent wave-front propagation then proceeded from the left and middle of the map and traveled downward and to the right to exit the scan region 57 ms after the shock onset. Note that continuous downward propagation did not occur at the right side of the map, as indicated by the closely spaced isochrones. Panel C shows optical traces from selected sites in this example. These recordings demonstrate the temporal and spatial sequence of myocardial membrane voltage changes accompanying the spread of activation along a line down from the base to the apex. Activation times immediately before and after the shock are given respectively on the left and right sides of the vertical shock marker. Trace a is taken from a site at the top of the scan region. It shows that a fibrillation action potential upstroke had already taken place 4 ms before the onset of the shock and that the shock had no effect on the membrane potential at this site. Trace b is taken from a site within the 10-ms isochronal contour and shows early generation of an action potential (CI, 81% [or 1.6 SD] below VFCL; V_m, 20% APA) coincident with the application of the shock. Trace c was taken just outside of the 10-ms isochrone, and it displays a premature depolarization (CI, 74.8% [or 2.1 SD] below VFCL) coinciding with the shock, which this time was smaller in amplitude and arising from a more depolarized takeoff potential (V_m, 24% APA). The depolarizations marked at 7 and 12 ms in traces b and c both meet the criteria for shock-induced depolarizations as described above. In contrast, optical recordings d and e, taken from more apical sites, show no shock-induced depolarization but instead depict interrupted (trace d) or continued (trace e) repolarization of the preceding preshock action potential. From the isochronal map and inspection of the optical recordings, we see that the immediate postshock propagating activity in the scanned region (ie, activations marked at 31 and 46 ms in traces d and e) originated from the area of shock-induced depolarization. In Fig 5, panel D is a representation of the same activation map as shown in panel B but without the individual activation times. The hatched areas encompass all sites that displayed shock-induced depolarization according to our definition; we call this the shock-depolarized area. Further, we also designate a limited perimeter of the shock-depolarized area that is capable of giving rise to propagating activity as the BSDA. Sites constituting the BSDA are circled or boxed in panel D. The midventricular hatched band is located just ahead of the preshock wave front, an area, as indicated by the optical recordings, that was most recovered at the time of the shock. Instead of being effectively halted, the preshock wave front was "reset" by the shock. Our preliminary results have shown that postshock activity arises in this manner for all but trivial strength shocks.¹⁹ The purpose of the present investigation was to determine the factors responsible for or indicative of why postshock wave fronts propagate or fail to propagate away from areas of shock-induced depolarization, such as in this example.

View larger version (67K): [in this window] [in a new window]   Figure 5. Example of immediate postshock propagation of the shock-induced depolarization shown to illustrate the methods of analysis. This scan was taken from the right ventricle; VFCL was 76.7±9.4 ms. A, Isochronal map of preshock fibrillation activity. Activation times are given in milliseconds with respect to the time of the shock. Isochronal lines are drawn at 10-ms intervals. B, Map tracking activity before, during, and after delivery of a shock at time zero. There is a wide area of near-synchronous depolarization ahead of the preshock wave front, from which propagated activity ensues on the left side but partially blocks on the right. C, Optical recordings obtained from circled sites, labeled a through e in panel A. Activation times are given next to the action potential upstrokes, and the time and duration of the shock are given by parallel vertical lines through all traces. Optical recordings show no shock effect at site a, shock-induced depolarizations at sites b and c, and propagated action potentials launched by the shock response at sites d and e. D, Map from panel B without activation times. Areas of shock-induced depolarization are shaded, as they are in subsequent maps. Sites forming the BSDA are indicated by circles at those sites where postshock propagation originated and by boxes where wave-front block occurred instead.

To consistently identify BSDA sites, only those sites displaying shock-induced depolarization located in the most depolarized myocardium at the time of the shock, facing an increasing gradient of depolarization (ie, the repolarizing wake of the prior wave front), were selected. For example, consideration of the pattern and direction of the preshock wave-front propagation in panel A of Fig 5 predicts that at the time of the shock, a gradient of increasing depolarization will exist from the leading edge of the next VF wave front (panel B, isochrone 0) down to the apex. Accordingly, BSDA sites were only considered along the lower border of the midventricular hatched band and not along the upper border of this region or the upper border of the apical hatched zone (although propagation against a repolarizing wake is theoretically possible, no instances of such were observed). Immediate postshock propagating activity never arose from sites other than BSDA sites identified by our criteria. Sites adjacent to the BSDA, in a more depolarized state than BSDA sites at the time of the shock, were not directly depolarized as a result of the shock (eg, compare responses to the shock in traces c and d of panel C). Maps were then analyzed to determine whether or not propagation ensued from each BSDA site. Propagation was said to occur when propagated action potentials were recorded at neighboring sites not directly depolarized by the shock (eg, the activations at 31 and 46 ms in traces d and e of panel C). Block was said to occur at these sites when no such activations arose within 20 ms of the corresponding BSDA site activation time. For the intersite distances typical of our experiments, this represented conduction velocities of <0.1 m/s, which corresponds to the lower limit of conduction velocities reported for normal ventricular myocardium.^{42 43 44} Some sites (22 of 898) analyzed in this study met the criteria for postshock propagation in one direction and block in another direction and, hence, were included in both groups. Panel D illustrates the concepts of propagation and "block" at BSDA sites. BSDA sites are located at the lower edge of the hatched shock-depolarized area and are outlined by either a circle or a box. Propagation ensued from the circled sites, whereas block occurred at the boxed sites. Because different sites constituting a single border can propagate or block, as in this example, and because one episode may show more than one BSDA in the scan region at shock time, it was possible for a single episode to show both propagation and block at the BSDA. Although the criteria for shock-induced depolarization were numerically arbitrary, reasonable alternate definitions would only slightly change the position of the BSDA from where we define it and lead to the same general results we present here.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
We analyzed a total of 331 episodes of attempted defibrillation in 23 hearts. Of these, 200 NDF and 73 DF episodes in 21 hearts were in the presence of 2 µmol/L D600, and 58 episodes (36 NDF and 22 DF) in four hearts were without this contraction-suppression agent (including data from two hearts before the addition of D600). A total of 795 and 103 BSDA sites (NDF and DF episodes, respectively) were identified and analyzed in the D600 data group, and 79 NDF and 58 DF BSDA sites were identified in the non-D600 data group. Results from the two groups are presented separately.

Occurrence of Propagation From the BSDA
Fig 6 shows an example of postshock propagation from the BSDA. Unlike the example in Fig 5, in this and the following example, the shock was applied before a preshock VF wave front entered the scan area. Panel A of Fig 6 is an activation map of the last wave front in the scan area before application of a 150-V (dialed voltage) shock, which failed to defibrillate. The map shows a single fibrillation wave front that propagated without interruption from apex (bottom of map) to base (top) during the time interval 80 to 37 ms before the shock. The map in panel B depicts the subsequent activity in the scanned region during and after the shock. As determined by the optical recordings, the directly shock-depolarized area encompassed the hatched zone at the bottom of the map; the BSDA was the upper border of this hatched area. The map depicts the path taken by the propagated activity immediately after the shock. Activation wave fronts propagated upward from the left and right sides of the BSDA (isochrones 20 and 30), merged above a small area of conduction block (indicated by the thick line) in the middle of the BSDA (30 to 40 ms), and exited at the basal edge of the scanned region 40 to 45 ms after the shock. Panel C shows optical recordings taken from selected sites (circled on the maps in panels A and B). Activation times are given to the right of each action potential upstroke. The traces are arranged in the apex-to-base direction of the preshock wave-front propagation. Recording sites a through c were located within the shock-depolarized area. Depolarizations at these sites were shock-induced according to our criteria, and as in Fig 5, displayed decreasing upstroke velocity and amplitude with decreasing shock CI and increasing proximity to the BSDA. Trace c is from a site along the BSDA itself. At this site, the shock CI was 60 ms, or 76.4% VFCL, and the V_m was 39% APA. The average CI and V_m values for all the sites constituting the BSDA in this example were 76±3% and 33±15%, respectively. Trace d was taken from a site just beyond the shock-depolarized area. Immediately after the shock, there was a slowly developing small depolarization that, according to our criteria, was not directly shock-induced but that may represent electrotonic spread from the shock-induced active depolarization at nearby sites. Subsequent regenerative activation occurred 25 ms after the shock, and examination of the map in panel B shows that its likely origin was the BSDA. Trace e, taken from a site more distal to the BSDA, shows uninterrupted repolarization of the preshock action potential, followed by an action potential propagated from the shock-depolarized area 32 ms after the shock.

View larger version (69K): [in this window] [in a new window]   Figure 6. Example of propagation from the BSDA without a preshock wave front in view at the time of the shock. This scan was taken from the left ventricle; VFCL was 86±2.1 ms. A, Preshock fibrillation activation pattern is shown. B, Postshock propagation originates from the left and right of the shaded region of shock-induced depolarization, whereas the wave front blocks in the center. C, Optical tracings from selected sites demonstrate direct excitation by the shock at sites a and b, which gave rise to the propagated wave front represented by activations at 14, 25, and 32 ms in traces c through e, respectively. The dialed voltage was 150 V.

In contrast to Fig 6, Fig 7 shows an example in which no propagated depolarization arose from the BSDA. The example was taken from the same experiment as Fig 6 but from an episode that used a stronger shock of 240 V dialed on the defibrillator. Panel A shows a roughly apex-to-base course of a preshock fibrillation wave front, approximately similar to the pattern in Fig 6A. As in the previous example, a shock was applied before a subsequent VF wave front entered the scan region. Fig 7B shows the postshock pattern of activation. The only sites that showed depolarization up to 50 ms after the shock are found within the area of shock-induced depolarization, highlighted by the hatching near the apex. In contrast to the previous example, there was no continued postshock propagating activity arising from the BSDA. After the shock-induced depolarization marked at 10 ms, there was a period of electrical quiescence (ie, no sites exhibited membrane depolarization) until an activation broke through in the upper left quadrant of the scan area 57 ms after the shock (not shown). Fig 7C shows optical recordings obtained at the same five sites as in Fig 6C, similarly arranged from apex to base. Traces a through c are taken from sites within the hatched area and show shock-evoked active depolarizing responses from well-depolarized myocardium. (We^{27 28} and others¹⁴ have previously shown that sufficiently strong shocks are able to elicit an additional phase of depolarization from already depolarized myocardium.) In this example, there is no clear relation between the rate and amplitude of the shock-induced depolarization and proximity to the BSDA, probably because the preshock pattern of activation (Fig 6A) was such that the shock CI was essentially identical (40 ms) at each of these sites. Site c is located on the BSDA. The shock CI at this site was 40 ms, or 57% VFCL, and the V_m of the shock response was 66% APA. The average CI and V_m values for all BSDA sites were 54±4% VFCL and 66±6% APA. Trace d shows only interrupted repolarization as a result of the shock, and trace e shows no effect of the shock on the repolarization of the preshock action potential. The first postshock action potential at these sites was unrelated to the shock responses evoked at the bottom of the scan area, consistent with the fact that propagation failed at the BSDA.

View larger version (60K): [in this window] [in a new window]   Figure 7. Example of complete block at the BSDA, from the same sites as illustrated in Fig 6. A, Preshock fibrillation activation pattern is shown; VFCL was 70.8±9 ms. B, Shock-induced depolarization occurred in the shaded region at the bottom of the map, with no subsequent propagation. C, Optical traces show shock-induced depolarization arising from high levels of membrane potential in traces a through c. However, traces from sites d and e show its failure to propagate into regions unaffected by shock. Subsequent activations invaded the region under study from elsewhere. The dialed voltage was 240 V.

Without mapping the entire heart, it is impossible to ascribe with certainty each episode of failed defibrillation to particular observed postshock wave fronts. However, their presence undoubtedly favors rather than hinders continued fibrillation. We observed wave-front propagation from at least one BSDA site in 30% (60 of 200) of all NDF episodes. If postshock wave fronts are responsible for failed defibrillation, it is possible that these arose in the remaining 70% of NDF episodes in areas not under view by the scanner. Consistent with this conjecture is the fact that an area approximately one third of the total ventricular epicardial surface was scanned in this study. In contrast to the failed defibrillation shocks, postshock propagation was observed in only 5% (4 of 73) of all successful defibrillation episodes; moreover, this propagation was generally short-lived (eg, <20-ms duration). Hence, although postshock activity was six times less frequent in successful defibrillation, its presence did not preclude defibrillation.

Influence of Shock CI and V_m on Postshock Propagation From the BSDA
Mapping results such as those presented above led us to hypothesize that the likelihood of wave-front propagation from the BSDA was dependent on the state of myocardial refractoriness prevailing at the BSDA at the time of the shock. For example, although the BSDAs appeared at similar locations in the same heart in the separate examples illustrated in Figs 6 and 7, the absence of postshock propagating activity in Fig 7 may be because the latter BSDA arose earlier and in more depolarized and, hence, more refractory myocardium than the former BSDA. To test this possibility, we measured the shock CI and V_m of the shock response at all BSDA recording sites and correlated the measurements to the incidence of wave-front propagation from these sites. To compare the effects of shocks against the unperturbed incidence of wave-front block during ongoing VF, we applied our analyses to VF episodes subjected to a sham, or 0-V, shock. The data set for this analysis consisted of a subset of 44 VF episodes, with the sham shock considered being delivered one fibrillation cycle length (as determined for that episode) before the real shock. This sample yielded 213 sham BSDA sites for analysis, or 20% as many real BSDA sites analyzed per heart. BSDA sites were present within the scan area in 33 of the 44 sham episodes. Since our BSDA detection criteria located depolarizations temporally related to the shock, application of the same algorithm to sham shocks identified the leading edges of native VF wave fronts.

Fig 8 plots the incidence of propagation from BSDA sites on the ordinate versus shock CI on the abscissa. Solid circles are data obtained from 898 individual BSDA sites resulting from real shocks; open circles denote control data obtained from 213 sites after sham (0-V) shocks. Solid lines are best fits through the data using a four-parameter logistic equation as in Fig 3; broken lines represent 95% confidence intervals of the fits. Little or no propagation took place when the CI was <50% VFCL for real shocks. Incidence of propagation then increased from 5% to 86% with increasing CI between 50% to 100% VFCL. The most straightforward interpretation of these data is that when the BSDA arose at short CIs (<50%), the shock-induced depolarization adjoined myocardium whose excitability was still unrecovered from previous activation and was therefore incapable of supporting action potential propagation. BSDAs elicited at longer CIs, after the effective refractory period, were then increasingly likely to propagate into myocardium adjoining the BSDA. The open circles show the dependence of propagation incidence on CI at the sham BSDA sites. In contrast to real shocks, BSDAs for sham shocks were never detected at CIs of <50% VFCL. This suggests that approximately half the VFCL constitutes the lower limit of native VF cycle lengths and, hence, the effective refractory period. The presence of solid circles at these shorter CIs reflects the ability of real shocks to elicit depolarization during this period. Propagation incidence increased sharply between 50% and 80% CI, followed by a plateau phase at 89% incidence. The curve for real shocks is shifted to the right with respect to the sham shock curve (confidence intervals do not overlap except at the longest coupling intervals). At nearly every CI, leading edges of depolarization were less likely to propagate after a real shock than during free-running VF. This implies that in addition to advancing the leading edge of depolarization into more refractory myocardium, shocks may have one or more additional effects that promote the failure of this depolarization to propagate.

View larger version (21K): [in this window] [in a new window]   Figure 8. Incidence of propagation from BSDA sites as a function of CI at those sites, for real () and sham () shocks. In this and subsequent figures, propagation incidence data were fitted by four-parameter logistic (same equation as in Fig 3) regression (solid lines, R2=.989 and .992 for real and sham shocks, respectively). The 95% confidence intervals are shown by the dashed and dotted lines for real and sham shocks. Real shocks at short CIs (<50% VFCL) rarely gave rise to propagating wave fronts. For sham shocks, no BSDAs (spontaneous wave fronts) occurred in this range of CIs, presumably because of myocardial refractoriness at short CIs. Note that the 95% confidence intervals overlap only at the longest CIs.

Because VF action potentials vary considerably in both their amplitudes and durations, we examined how reliably the CI by itself can predict refractoriness in VF. We plotted the measured V_m as a function of CI for all BSDA sites for which V_ms were available. V_ms were unavailable in some episodes because a rescue shock failed to promptly defibrillate and therefore prevented signal calibration by a control action potential. Also, baseline shift induced by the shock at some sites precluded the accurate determination of their optical resting potential level. Fig 9 shows a clear trend of decreasing V_m with increasing CI, as would be expected, but there is considerable scatter around this trend. (Linear regression analysis [SigmaPlot] yielded a slope of -0.59 and R²=.52.) At late CIs (eg, 80% VFCL), we find V_ms from near resting levels to over 70% APA, with most measurements within the range of 10% to 50% APA (35.9±11.8% mean±SD). Because of this loose correlation between measured CI and V_m and since the optical recording technique could provide an indication of membrane voltage levels, we repeated the analysis of propagation incidence at BSDA sites as a function of V_m, believing that this would provide a more direct and more precise measure of local refractoriness.

View larger version (43K): [in this window] [in a new window]   Figure 9. Vm as a function of CI for 757 BSDA sites. There is considerable scatter in the plot at any fixed CI, probably reflecting beat-to-beat variations in APA and duration during VF.

Fig 10 depicts the relationship between propagation incidence and V_m at 757 BSDA sites with real shocks and 175 BSDA sites with sham shocks. The data are fit by a four-parameter logistic equation. The real-shock data (solid circles) follow a negative sigmoid relationship, with a maximum asymptote at 75% propagation at low V_m and the minimum asymptote at 0% propagation (100% block) reached near 65% APA. The negative slope is in keeping with the hypothesis that propagation block should increase with increasing refractoriness and, hence, increasing depolarization, prevailing at the BSDA at the time of the shock. The curve for sham shocks shows a steeper dependence on V_m over midrange potentials. It contains no data points at potentials beyond 62.5% APA (the largest value in the 60% to 70% voltage bin), indicating that spontaneous activations in VF never arose from levels more depolarized than this. Another difference between sham and real shocks is that propagation was never blocked (100% incidence) when BSDA sites occurred at low membrane voltages (30% APA) in free-running VF, whereas its incidence was only 75% despite comparable V_m for real shocks. This result suggests that defibrillation shocks have effects other than advancing the leading edges of depolarization into more refractory areas, which favor the cessation of wave-front propagation. Further supporting this notion, BSDAs arising at any V_m (0% to 70% APA) had lower incidences of propagation in the event of a real shock, compared with sham shock. This is reflected as a shift to the left of the curve for real shocks with respect to that for sham shocks. The 95% confidence interval curves are nonoverlapping, except when propagation incidence is zero. That both curves reached 0% propagation at the same point, well before 100% APA, is in keeping with the finding that comparable levels of depolarization marked the transition between the relatively refractory and still excitable phase of the cardiac cycle, and the effective refractory period.^{15 45}

View larger version (20K): [in this window] [in a new window]   Figure 10. Incidence of propagation from BSDA sites as a function of Vm for real () and sham () shocks. Logistic regression was performed as in Fig 8, with R2=.989 and .986 for real and sham shocks, respectively. BSDAs arising at repolarized levels (low Vm) reached 100% propagation incidence for sham shocks but only 75% for real shocks. There is a reduction in propagation incidence with increasing depolarization at the BSDA. Propagation incidence was 0% for Vm >60% APA for real shocks, whereas no VF activations, blocked or propagated, arose from Vms more depolarized than 62.5% APA with sham shocks. Propagation incidence after real shocks was less than after sham shocks in all but one Vm bin. The 95% confidence intervals overlap only at propagation incidences near zero.

Lack of Effect of D600
We investigated the possibility that the Ca²⁺ channel blocker D600 could qualitatively or quantitatively affect the incidence of propagation and block of depolarization arising at the BSDA. Accordingly, we recorded 58 attempted defibrillation episodes in four hearts in the absence of D600. Because cardiac contraction following the rescue shock prevented accurate determination of the baseline fluorescence corresponding to resting membrane potential, we were unable to make optical measurements of membrane potential. Therefore, we analyzed these data to determine propagation outcome at the BSDA as a function of shock CI only. Fig 11 shows propagation incidence as a function of shock CI for episodes without D600 (open circles). Results with D600 are reproduced for comparison (solid circles). The plot for no D600 includes data from 137 BSDA sites obtained from 19 episodes that yielded BSDAs. As in the presence of D600, there was no propagation from the area of shock-induced depolarization when the CI at its border was <50% VFCL. The logistic fit follows a curve similar to that for D600, and the 95% CIs overlap at all CIs except for a window of CIs between 70% and 85% VFCL, where they lie very close. Hence, the curves for the two data sets are scarcely distinguishable. D600 caused a slight lengthening of the mean VFCL (73.8±6.85 ms [n=118] versus 66.3±7.9 ms [n=19], average calculated VFCL for all D600 and no D600 episodes, respectively; P<.05).

View larger version (21K): [in this window] [in a new window]   Figure 11. Propagation incidence as a function of CI for 137 BSDA sites in the absence of D600 (). Data from shocks with D600 () are reproduced for comparison. Logistic regression was performed as in Fig 8, with R2=.957 and .989 for no D600 and D600 episodes, respectively. The curve for no D600 is similar to that for D600, indicating that the use of D600 did not significantly affect the dependence of propagation incidence on CI. Both curves show increasing incidence of propagation with lengthening CI, beginning at 50% VF cycle length. Notably, the 95% confidence interval for the no-D600 curve (dotted lines) overlaps that for the D600 curve (dashed line) across most of the CI range. Data points for shocks without D600 show less smooth distribution and attain 100% propagation incidence, probably because of the smaller number of observations in this set.

Shock-Induced Reentry
Recently, the de novo induction of reentrant circuits by shocks has been advanced as a mechanism for failed defibrillation.^{7 8 10} The ULV hypothesis states that shocks stronger than 1 J terminate all VF wave fronts,²⁴ after which the shock reinitiates fibrillation in the same manner as it can be initiated by stimulation during the vulnerable period.^{7 46} The mechanism, first proposed by Winfree,⁴⁷ consists of the formation of a "phase singularity" or "critical point" in a plane where critical gradients in shock field strength and tissue refractoriness orthogonally intersect.

In 8 of the 331 shock episodes constituting both parts of this study (273 with and 58 without D600), a full reentrant circuit was observed within the mapped region immediately after application of the shock. In three of these, the postshock reentry could be attributable to the formation of a critical point on the BSDA, since wave-front propagation and (unidirectional) block occurred on either side of the critical point. In the other five instances of postshock reentry, the origin of the wave front, the functional line of block, or both could not be attributed to any measurable effect of the shock (ie, membrane depolarization), and a critical point as described above was not present. Fig 12 shows an example of critical point formation and induction of reentry after a failed defibrillation attempt. Panel A is an isochronal map of a preshock VF wave front that entered the scan region at the left 86 ms before a 140-V (dialed voltage) shock and traveled uniformly toward the right to exit the scan area at -32 ms. This pattern predicts a left-to-right gradient in repolarization and refractoriness. Panel B shows the subsequent VF wave-front curling around a zone of block (thick line at the left) and invading roughly the left third of the scan region by the time of the shock (isochrone 0). The crosshatched region depicts the area of myocardium directly depolarized by the shock; the BSDA is the right border of this zone. (Where two activation times are given for a single given site, the lower italicized number represents a second activation recorded at that site within the time frame of the map.^{25 35} ) No propagation emanated from the top portion of the BSDA; the resulting conduction block is depicted by the thick line. However, below a point along the BSDA, a wave of activation did conduct away from the border of shock-depolarized myocardium. It propagated rightward and upward, around the bottom of the line of block to continue counterclockwise and to eventually cross the original line of block at 53 ms after the shock and reactivate myocardium that had been previously depolarized by the shock and had since recovered. Although this map extends only so far as to show activity until 60 ms after shock, the reentrant wave front continued to rapidly activate the area hatched in the figure and was conducted from the bottom center of the map upward and rightward to ultimately block at the location of the 30-ms isochrone. In Fig 12, panel C shows optical recordings taken from seven sites during this episode. Trace a shows the shock arriving just after a fibrillation upstroke on the far left of the scan region and exhibits no membrane response. Site b is located within the hatched area of panel B and shows shock-induced depolarization. Traces c through f display no direct effect of the shock on local membrane potential but document the path of the reentrant impulse launched from the lower part of the BSDA through the increasing latency of the depolarizing deflections. In trace d, note that the propagated action potential, marked at 31 ms, was slowly developed and was low in amplitude and that it corresponded, on the map, to the site of slowed conduction velocity. In trace g, activation is seen at 57 ms as the propagated wave returned to invade the region previously directly depolarized by the shock. The last action potentials of traces c and d (marked at 78 and 80 ms after shock, not shown) represent the continuation of the reentrant part of the propagated wave, which eventually blocked at the site of recording e (isochrone 30 on the map of panel B).

View larger version (71K): [in this window] [in a new window]   Figure 12. Example of wave front reentry formed de novo by a shock. This scan was taken from the right ventricle; VFCL was 70.1±5.8 ms. A, Isochronal map of preshock fibrillation activation pattern. A fibrillation wave-front conducted across the scan region from left to right. B, Map showing subsequent preshock activity on the left and a shaded area of shock-induced depolarization in the center. This example meets the criteria for reentry by critical point formation. C, Optical recordings from sites labeled a through g in panel A. See "Results" for complete description of the events depicted in this figure.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The application of electric shocks to the heart is currently the only reliable means to terminate VF. By extension of what is known about atrial fibrillation,^{34 48} VF is widely considered to consist of multiple reentrant wave fronts undertaking wandering paths.^{1 2} Given this reentrant self-sustaining nature of fibrillation, elimination of propagating electrical activity in the heart would guarantee defibrillation (barring shock-induced ectopic impulses, which could cause defibrillation to effectively fail). Observations by Wiggers⁴⁹ in 1940 led him to state that all vestiges of electrical activity had to be abolished for defibrillation to be successful; however, it was shown later that defibrillation could also be achieved even when propagated activity persisted after the shock.^{5 6 24 26} Since the presence or absence of postshock activity plays a key role in determining whether defibrillation ultimately succeeds, any adequate understanding of the defibrillation process must include a description of how shocks eliminate propagated activity. Although wave-front termination forcibly plays a central role in defibrillation, other shock effects not considered in the present study (eg, repolarization resynchronization^{28 50} ) may also influence shock outcome, particularly in the setting of persisting propagation.

The simplest explanation for the presence of postshock activations is that they represent the continuation of preshock fibrillating wave fronts that survived the shock.²⁶ However, on the basis of low-resolution electrical mapping, Shibata et al⁵¹ speculated that in some cases a shock could "reset" ongoing VF wave fronts, so that the postshock impulses do not represent preshock wave fronts but rather originate from the border of areas directly activated by the shock. In the present study, we applied an optical mapping technique to ascertain the origin of postshock propagating activity and to identify and quantify the mechanisms that determine whether postshock wave fronts propagate or not. Our results confirm Shibata's supposition that postshock wave fronts can originate from the BSDA in VF. We have found that the likelihood of propagation of such wave fronts diminishes with increasing refractoriness at the BSDA and that this propagation is completely suppressed when BSDAs arise in effectively refractory myocardium. Our results also point to at least one additional mechanism by which shocks promote wave-front block.

Postshock Propagating Activity Arises From the BSDA
Taken together, the activation maps and optical recordings obtained in the present study unequivocally demonstrate that postshock propagating wave fronts can arise from the border of myocardium directly activated by a shock (Figs 5 and 6). In those cases in which a preshock fibrillation wave front was observable at the time of the shock, the shock appeared to have advanced the wave front (eg, see Fig 5), a process we have described as "resetting."¹⁹ (We used the term "reset" only to describe, by analogy with the advancement of the circulating wave front and subsequent reset of ventricular tachycardia,^{52 53} the events that immediately follow a shock.) However, such a reset is not the general outcome, since VF does not consist of a single, stable, and predictable circuit and because postshock activation may follow pathways that might not have been taken had the shock not been applied (eg, Fig 2). Consequently, we studied postshock impulse conduction and block at all BSDA sites, without regard to whether a preshock wave front was observable at the time of the shock (eg, Figs 6 and 7).

Excitable Gap in VF
The fact that it is possible for a shock to depolarize a considerable area of myocardium ahead of ongoing VF wave fronts and yet observe continued propagating activity from its border implies the existence of a significant excitable gap in this preparation to global shocks during VF. The incidence of propagation from BSDA sites was near zero when the CIs at those sites were shorter than 50% of the VFCL and extrapolation of the curve for sham shocks to 0% propagation yielded a similar result (Fig 8), suggesting 50% VFCL as an upper limit for the excitable gap in VF. In a recent study of spontaneous termination of reentry during VF in dogs, Cha et al⁵⁴ found refractory periods corresponding to an excitable gap of 39% to 61% VFCL. Although our results are in agreement with those of Cha et al, an important difference in the present study is the delivery of shocks that, as discussed below, could affect refractoriness ahead of the shock-induced wave fronts. The existence of an excitable gap implies that it should be possible to locally entrain VF by pacing, as has been done for atrial fibrillation.⁵⁵ This latter study⁵⁵ showed that stable regional entrainment was possible at stimulation cycle lengths shorter than 70% of all fibrillation intervals, indicating the existence of an excitable gap during a majority of atrial fibrillation activation intervals. In a recent publication, KenKnight et al⁵⁶ have demonstrated the presence of an excitable gap in VF by achieving regional capture by applying strong pacing stimuli during VF in a swine model.

Cessation of Postshock Fibrillating Activity
Possible propagation of shock-induced depolarizations in attempted defibrillation means that it is insufficient to simply stimulate the most excitable myocardium immediately ahead of VF wave fronts and cause their collision into prematurely refractory tissue to halt propagation. A key objective, recognized by Shibata et al,⁵¹ becomes the determination of how wave-front resetting is prevented to achieve postshock termination of propagated activity and, ultimately, defibrillation. Data summarized in Figs 8 and 10 show that directly depolarizing larger areas by the shock, thereby advancing the leading edge of the shock depolarized area (BSDA) into less repolarized, more refractory myocardium, results in a decreasing probability of propagation from the BSDA and increasing likelihood of wave-front block, ie, failure to conduct away from these sites. Prevention of postshock propagation is ensured when the shock is sufficiently strong to evoke depolarization from V_ms already depolarized to 60% of a control action potential. Since the majority of fast Na⁺ channels are inactivated in this voltage range,⁵⁷ the BSDA includes what is classically considered absolutely refractory myocardium.⁵⁸ Since strong enough shocks can elicit regenerative responses from all physiological levels of transmembrane voltage,²⁷ we refer specifically to a state of effective refractoriness, from which propagating responses cannot be evoked.⁵⁸ In such cases, the entire excitable gap is depolarized, so that the new leading edge of depolarization abuts only myocardium that is also in the effective refractory period, thereby precluding propagation.

The simple scheme presented above explains the near-zero incidence of postshock propagation when the BSDA is located in effectively refractory tissue; however, it does not fully account for the different propagation incidences seen at lower V_m and longer CI. This is to be expected for a number of reasons. The transition from excitability to complete refractoriness is not instantaneous. Over a finite portion of the cardiac cycle (the relative refractory period), increasingly strong sources of depolarizing current are required to excite cells at shorter CIs and more depolarized potential.¹⁵ Consequently, with increasing relative refractoriness at the BSDA and all other factors being equal, a decreasing proportion of BSDA sites will provide sufficient depolarizing current to bring the adjoining myocardium to threshold and initiate propagation. In this respect, continued postshock propagation remains primarily a function of the V_m at the BSDA. This V_m is ultimately determined through the interaction of local shock strength and the distribution of repolarization (refractoriness) gradients caused by preshock VF activations at the time of the shock. Other factors, not explicitly considered in the present study, almost certainly play a role in determining wave-front propagation at the BSDA. These factors might serve to modulate the effect of V_m at the BSDA, but then only when the cells at these sites are no longer in their effective refractory period at the time of the shock.

Shocks may affect the excitability of cells beyond the BSDA without actually depolarizing the membrane potential. This may explain the fact that for BSDAs located in similarly refractory myocardium (as indicated by CI or V_m in Figs 8 and 10), real shocks always resulted in lower propagation incidences than did sham shocks. This difference indicated that shocks had some effect that promoted conduction block at the BSDA and that was not entirely due to advancing the leading edge of depolarization into more refractory myocardium. Prolongation of the refractory period^{59 60 61} can take place just beyond the BSDA, where membrane potential can be maintained and repolarization delayed (eg, trace d of Figs 5C and 7C). Such responses are distinct from graded responses,¹⁵ since they show no depolarization. Nevertheless, these nondepolarizing responses could block impulses that would otherwise be conducted away from the BSDA by producing additional refractoriness both in time (prolongation) and space (extension ahead of the BSDA).

Other modulating factors that may determine the fate of shock-induced depolarizations are not directly V_m dependent, such as the effect of the orientation of the BSDA with respect to the major (longitudinal) axis of the cardiac fibers. Uniform anisotropy has been shown to significantly influence the margin of safety for impulse conduction in partially uncoupled myocardium⁶² and myocardium with impaired excitability^{21 63} (but see Spach et al⁶⁴ ). The latter condition certainly is the case in VF, since resting membrane voltage levels are essentially never attained.^{29 30 31 32} The orientation of a shock field relative to myocardial fiber orientation has been reported to affect certain shock-response parameters, such as the shock voltage gradient threshold for eliciting excitation^{14 65} and action potential prolongation.¹⁴ Hence, it is likely that anisotropy can directly affect BSDA shape and location in addition to its propagation outcome. Another potentially modulating factor is the curvature of the BSDA in space. A concave wave-front supplies more current and, hence, is more likely to depolarize to threshold and activate a given unit of adjacent myocardium compared with a convex wave front.^{66 67} Since the BSDAs in the present study varied in curvature and were presumably formed at many different angles with respect to myocardial cell axes, these factors may mask an even stronger dependence of postshock propagation on BSDA V_m than we have shown here.

Knisley et al¹⁴ invoked, in addition to absolute membrane potential, a role for postshock spatial gradients in membrane potential in determining wave-front propagation or block in which a steep gradient would predispose to postshock propagation. We have not specifically evaluated this parameter in the present study, nor have we considered the possibility of propagation against an increasing gradient in repolarization. However, the Knisley study predicts steep postshock V_m gradients to form only at CIs over 95% of a paced action potential duration. Since membrane potentials so near to resting level are rarely attained in VF, their results are difficult to extrapolate to actual defibrillation.

Limitations
The fact that the present study was carried out in an isolated continuously perfused rabbit heart model limits to what degree the results presented herein can be extrapolated to the in situ ischemic condition of clinical fibrillation. Nevertheless, previous work from this laboratory has established that defibrillation shocks produce qualitatively similar electrophysiological effects during ischemic VF in the rabbit heart²⁸ and in the in situ swine heart,³³ as in the present model.

To determine whether the use of the Ca²⁺ channel blocker D600 could substantially affect the dependence of propagation incidence on refractoriness at the BSDA, the experimental protocol was executed in the absence of D600 in four hearts. As described in "Results," the propagation incidences plotted as a function of shock CI showed only minor differences with the plot obtained in the presence of 2 µmol/L D600. In both cases, nearly all of the change occurred between 40% and 80% VFCL, when major change in excitability occurs. Although any minor differences might be related to D600, they may also reflect variation due to the imprecision of CI in estimating V_m in VF (eg, see Fig 9) and, therefore, the prevailing level of refractoriness.

Correlation With Electrical Mapping Studies
The results of the present study complement several previous studies that correlated the presence of propagating activity after the shock with failed defibrillation.^{24 26 51} Our detection by optical mapping of propagated electrical activity immediately following a shock is consistent with the conclusion of Witkowski et al,²⁶ who determined, on the basis of postshock activation times obtained using unipolar DC-coupled Ag/AgCl electrodes, that electrical activity does not cease and then recommence. However, our observation that postshock activity arose from the BSDA and is therefore due to the shock itself disagrees with another conclusion of that study, namely, that postshock propagation represents the continuation of preshock fibrillating activity. Identification of the origin of postshock wave fronts confirms one component of the ULV hypothesis, namely, that shocks can generate as well as terminate fibrillation wave fronts.¹¹ Two electrical mapping studies by Ideker and colleagues^{25 36} have shown unidirectional conduction after a shock given during VF. It was hypothesized that the postshock wave fronts might originate from the BSDA, as we have shown in the present study. In those studies, however, areas of shock-induced depolarization had to be surmised, because extracellular electrogram recording cannot detect stationary activation. If a shock-induced depolarization did not give rise to a propagated wave front, it would be missed. If a postshock wave front did conduct away from a BSDA, a finite "isoelectric window" would be observed because of gain switching and settling time and possibly because slowly developing low-amplitude activations occurring near the BSDA (eg, trace d, Fig 12C) were not detected. Daubert et al have shown that impulses can conduct from the BSDA for shocks delivered during paced rhythm, using a single monophasic action potential recording site to supplement electrical mapping.¹³

Electrical mapping studies of defibrillation have also reported a different type of early postshock activity, propagating in a focal pattern.^{24 25 36} We have observed no focal conduction attributable to the effects of a shock. On rare occasions (n=7), however, we have observed focal-type activations occurring 48±6 ms after the shock in areas where the shock did not affect repolarization. This occurred in instances in which the BSDA failed to initiate propagating activity anywhere along its length and in which the earliest postshock propagated activity came from within the mapped region rather than from the edge. We believe these "foci" were analogous to the earliest postshock activity described by the electrical studies. If one considers these foci to be the epicardial breakthrough of unaltered or shock-evoked wave fronts from elsewhere (eg, the septum), there is no need to invoke different mechanisms such as automaticity or microreentry to explain their presence. Breakthrough patterns would not necessarily be broad and synchronous as they are in regular rhythms because of prior fibrillation activity (ie, foci may be the earliest epicardial sites to regain their excitability). One study by Chen et al,²⁵ however, is more difficult to reconcile with this view, since it detected postshock focal activity (which sometimes repeated) using transmural (three-dimensional) mapping.

Implications for the Mechanisms of Defibrillation
The present results directly demonstrate that shocks themselves can be the origin of postshock propagating activity in defibrillation. In describing shock effects from a strictly biophysical, rather than functional, point of view, it can be said that in the process of shock-induced wave-front origination, preshock VF wave fronts are annihilated and are replaced by new wave fronts at the BSDA. However, there is continual electrical activity, before and after a shock, and therefore no real period of total electrical quiescence. Accordingly, we do not share the interpretation offered by Ideker and colleagues^{8 10 11 24 25 36} that failed defibrillation is tantamount to cessation and the subsequent regeneration, reinduction, or reinitiation of VF, because these terms imply that actual defibrillation took place. Although we recognize that fibrillation can be modified by a shock, we take defibrillation to mean, as it does clinically, ultimately successful termination (for more than a fraction of a second) of the fibrillation rhythm.

The ULV hypothesis specifically states that failed defibrillation occurs by the same mechanism that is responsible for VF induction by premature stimuli delivered during the vulnerable period of regular rhythm.^{7 8 24} This mechanism is induction of reentry by formation of a critical point.⁴⁶ Although the present study reports for the first time an instance of reentry induction by critical point formation in defibrillation (Fig 12), evidence suggests that such wave-front dynamics are not absolutely necessary for defibrillation to fail. Reentry by critical point was found in only 3 of the 236 NDF episodes (200 with and 36 without D600) analyzed in the present study. Critical points were not reported in electrical epicardial plaque mapping of 12 NDF episodes,²⁴ in a three-dimensional mapping study of 167 NDF episodes,²⁵ or in a high-resolution epicardial mapping study of 600 total shocks (number of NDF episodes were not reported).³⁶ The latter study did show examples of postshock reentry; however, these were described as leading circle reentry and could not be unequivocally attributed to critical point formation by the shock, since the area of shock-induced depolarization could not be directly determined because of limitations of electrical mapping. It may be that critical points formed outside the mapped region caused defibrillation to fail in the present study. However, this argument cannot be reconciled with the results of one electrical mapping study³⁶ that was designed to produce lower shock field strengths within the mapped area than elsewhere in the ventricle. (It has been established that earliest postshock activity occurs preferentially in areas of lowest shock field strength.^{9 51} ) The overall scarcity of critical points is not surprising, considering the requirement for approximately orthogonal alignment of shock field and repolarization gradients and the disorganized nature of VF wavelet conduction paths. Reentry can occur by mechanisms other than critical point formation, such as shock-induced reset of an existing reentrant circuit or delayed reentry of postshock wave fronts (Fig 8 of Reference 36; authors' unpublished data, 1993-1994). It may be that, like atrial fibrillation,³⁴ VF is not perpetuated by closed reentrant circuits but can persist through continued random reentry² of postshock propagating wave fronts, which may wander, collide, or split into daughter wavelets (eg, see Fig 2). Thus, we believe that the induction of a single closed reentrant circuit through the formation of a critical point represents a specific example among many possible ways defibrillation may fail as a result of postshock propagation arising from BSDAs in general.

Our results from DF episodes are consistent with previous investigations^{5 6 24 26} that found that it is not necessary to abolish all postshock propagating activity to achieve defibrillation. Work by Dillon²⁸ and others⁶⁶ suggests that shock-induced resynchronization of repolarization may allow for ultimate defibrillation, despite continued postshock propagation in some part of the heart, by thwarting the development of partial block and reentry or wave-front splitting. Resolving this issue will require further investigation.

	Selected Abbreviations and Acronyms

 

APA = action potential amplitude BSDA = border of the shock-depolarized area CI = coupling interval CM = critical mass DF, NDF episodes = defibrillating and nondefibrillating episodes ULV = upper limit of vulnerability VF = ventricular fibrillation VFCL = VF cycle length Vm = shock response takeoff potential

	Acknowledgments

 
This study was supported by the Sidney Kimmel Cardiovascular Research Center and grant R01 HL-49246 from the National Heart, Lung, and Blood Institute of the National Institutes of Health, Bethesda, Md, to Dr Dillon. Mr Kwaku was a recipient of a scholarship from the Fonds pour la Formation de Chercheurs et l'Aide a la Recherche, Quebec, Canada. The authors wish to express their sincere gratitude to Dr Rahul Mehra for insightful discussions.

	Footnotes

 
Previously published as preliminary results (Circulation. 1994;90[suppl I]:I-447; Pacing Clin Electrophysiol. 1995;18:808/1996;19:666).

Received October 26, 1995; accepted July 30, 1996.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Garrey WE. The nature of fibrillatory contraction in the heart: its relation to tissue mass and form. Am J Physiol. 1914;33:397-414.

2. Hoffman BF, Rosen MR. Cellular mechanisms for cardiac arrhythmias. Circulation. 1981;49:1-15.

3. Prevost J-L, Battelli F. La mort par les courants electriques: courants alternatifs a haute tension. J Physiol Pathol Gen. 1899;1:427-442.

4. Beck CS. Resuscitation for cardiac standstill and ventricular fibrillation occurring during operation. Am J Surg. 1941;54:273-279.

5. Mower MM, Mirowski M, Spear JF, Moore NE. Patterns of ventricular activity during catheter defibrillation. Circulation. 1974;49:858-861.[Abstract/Free Full Text]

6. Zipes DP, Fischer J, King RM, Nicoll AD, Jolly WW. Termination of ventricular fibrillation in dogs by depolarizing a critical amount of myocardium. Am J Cardiol. 1975;36:37-44.[Medline] [Order article via Infotrieve]

7. Chen P-S, Shibata N, Dixon EG, Martin RO, Ideker RE. Comparison of the defibrillation threshold and the upper limit of vulnerability. Circulation. 1986;73:1022-1028.[Abstract/Free Full Text]

8. Chen P-S, Wolf PD, Ideker RE. Mechanism of cardiac defibrillation: a different point of view. Circulation. 1991;84:913-919.[Free Full Text]

9. Witkowski FX, Kerber RE. Currently known mechanisms underlying direct current external and internal cardiac defibrillation. J Cardiovasc Electrophysiol. 1991;2:562-572.

10. Walcott GP, Walcott KT, Knisley SB, Zhou X, Ideker RE. Mechanisms of defibrillation for monophasic and biphasic waveforms. Pacing Clin Electrophysiol.. 1994;17:478-498.[Medline] [Order article via Infotrieve]

11. Ideker RE, Wolf PD, Tang ASL. Mechanisms of defibrillation. In: Tacker WA Jr, ed. Defibrillation of the Heart: ICDs, AEDs, and Manual. St Louis, Mo: CV Mosby Co; 1994:15-45.

12. Dillon SM. The electrophysiology of defibrillation: today and yesterday. In: Allesie MA, Fromer M, eds. Atrial and Ventricular Fibrillation Mechanisms and Device Therapy. Armonk, NY: Futura Publishing; 1996.

13. Daubert JP, Frazier DW, Wolf PD, Franz MR, Smith WM, Ideker RE. Response of relatively refractory canine myocardium to monophasic and biphasic shocks. Circulation. 1991;84:2522-2538.[Abstract/Free Full Text]

14. Knisley SB, Smith WM, Ideker RE. Effect of field stimulation on cellular repolarization in rabbit myocardium: implications for reentry induction. Circ Res. 1992;70:707-715.[Abstract/Free Full Text]

15. Kao CY, Hoffman BF. Graded and decremental response in heart muscle fibers. Am J Physiol. 1958;194:187-196.

16. Dillon S, Morad M. A new laser scanning system for measuring action potential propagation in the heart. Science. 1981;214:453-456.[Abstract/Free Full Text]

17. Hill BC, Courtney KR. Design of a multi-point laser scanned optical monitor of cardiac action potential propagation: application to microreentry in guinea pig atrium. Ann Biomed Eng. 1987;15:567-577.[Medline] [Order article via Infotrieve]

18. Dillon SM. Optical mapping. In: Shenasa M, Borggrefe M, Breithardt G, eds. Cardiac Mapping. Mt Kisco, NY: Futura Publishing; 1993:587-603.

19. Kwaku KF, Dillon SM. Stimulation of excitable myocardium not sufficient to halt fibrillation wavefronts. Circulation. 1994;90(pt 2):I-447. Abstract.

20. Kwaku KF, Dillon SM. Shock-induced depolarization of refractory myocardium prevents propagation of the defibrillation shock response. Pacing Clin Electrophysiol. 1995;18(pt 2):808. Abstract.

21. Schalij MJ. Anisotropic Conduction and Ventricular Tachycardia. Limburg, The Netherlands: University of Limburg; 1988. Dissertation.

22. Schalij MJ, Lammers WJEP, Rensma PL, Allesie MA. Anisotropic conduction and reentry in perfused epicardium of rabbit left ventricle. Am J Physiol. 1992;263:H1466-H1478.[Abstract/Free Full Text]

23. Gupta RK, Salzberg BM, Grinvald A, Cohen LB, Kamino K, Lesher S, Boyle MB, Waggoner AS, Wang CH. Improvements in optical methods for measuring rapid changes in membrane potential. J Membr Biol. 1981;58:123-137.[Medline] [Order article via Infotrieve]

24. Chen P-S, Shibata N, Wolf P, Dixon EG, Danieley ND, Sweeney MB, Smith WM, Ideker RE. Activation during ventricular defibrillation in open chest dogs: evidence of complete cessation and regeneration of ventricular fibrillation after unsuccessful shocks. J Clin Invest. 1986;77:810-823.

25. Chen P-S, Wolf PD, Melnick SD, Danieley ND, Smith WM, Ideker RE. Comparison of activation during ventricular fibrillation and following unsuccessful defibrillation shocks in open-chest dogs. Circ Res. 1990;66:1544-1560.[Abstract/Free Full Text]

26. Witkowski FX, Penkoske PA, Plonsey R. Mechanism of cardiac defibrillation in open-chest dogs using unipolar DC-coupled simultaneous activation and shock potential recordings. Circulation. 1990;82:244-260.[Abstract/Free Full Text]

27. Dillon SM. Optical recordings in the rabbit heart show that defibrillation strength shocks prolong the duration of depolarization and the refractory period. Circ Res. 1991;69:842-856.[Abstract/Free Full Text]

28. Dillon SM. Synchronized repolarization after defibrillation shocks: a possible component of the defibrillation process demonstrated by optical recordings in rabbit heart. Circulation. 1992;85:1865-1878.[Abstract/Free Full Text]

29. Surawicz B, Gettes LS, Ponce-Zumino A. Relation of vulnerability to ECG and action potential characteristics of premature beats. Am J Physiol. 1967;212:1519-1528.

30. Zhou X, Guse P, Wolf PD, Rollins DL, Smith WM, Ideker RE. Existence of both fast and slow channel activity during the early stages of ventricular fibrillation. Circ Res. 1992;70:773-786.[Abstract/Free Full Text]

31. Zhou X, Wolf PD, Rollins DL, Afework Y, Smith WM, Ideker RE. Effects of monophasic and biphasic shocks on action potentials during ventricular fibrillation in dogs. Circ Res. 1993;73:325-334.[Abstract/Free Full Text]

32. Downar E, Janse MJ, Durrer D. The effect of acute coronary artery occlusion on subepicardial transmembrane potentials in the intact porcine heart. Circulation. 1977;56:217-224.[Abstract/Free Full Text]

33. Menz V, Li KS, Schwartzman D, Dillon SM. In-situ optical action potential recordings of defibrillation in swine. Pacing Clin Electrophysiol. 1995;18(pt 2):877. Abstract.

34. Allessie MA, Lammers WJEP, Bonke FIM, Hollen J. Experimental evaluation of Moe's multiple wavelet hypothesis of atrial fibrillation. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology and Arrhythmias. Orlando, Fla: Grune & Stratton; 1985:265-275.

35. Chen P-S, Wolf PD, Dixon EG, Danieley ND, Frazier DW, Smith WM, Ideker RE. Mechanism of ventricular vulnerability to single premature stimuli in open-chest dogs. Circ Res. 1988;62:1191-1209.[Abstract/Free Full Text]

36. Zhou X, Daubert JP, Wolf PD, Smith WM, Ideker RE. Epicardial mapping of ventricular defibrillation with monophasic and biphasic shocks in dogs. Circ Res. 1993;72:145-160.[Abstract/Free Full Text]

37. Janse MJ, Van Capelle FJL, Morsink H, Kleber AG, Wilms-Schopman F, Cardinal R, D'Alnoncourt CN, Durrer D. Flow of `injury' current and patterns of excitation during early ventricular arrhythmias in acute regional myorcardial ischemia in isolated porcine and canine hearts: evidence for two different arrhythmogenic mechanisms. Circ Res. 1980;47:151-165.[Free Full Text]

38. Janse MJ, Kleber AG. Electrophysiological changes and ventricular arrhythmias in the early phase of regional myocardial ischemia. Circ Res. 1981;49:1069-1081.[Free Full Text]

39. Kimura J, Miyamae S, Noma A. Identification of sodium-calcium exchange current in single ventricular cells of guinea-pig. J Physiol (Lond). 1987;384:199-222.[Abstract/Free Full Text]

40. Levesque PC, Leblanc N, Hume JR. Role of reverse-mode Na⁺-Ca²⁺ exchange in excitation-contraction coupling in the heart. Ann N Y Acad Sci. 1991;639:386-397.[Medline] [Order article via Infotrieve]

41. Davy J-M, Fain ES, Dorian P, Winkle RA. The relationship between successful defibrillation and delivered energy in open-chest dogs: reappraisal of the `defibrillation threshold concept. Am Heart J. 1987;113:77-84.[Medline] [Order article via Infotrieve]

42. Scher AM, Young AC. Spread of excitation during premature ventricular systoles. Circ Res. 1955;3:535-542.[Abstract/Free Full Text]

43. Sano T, Takayama N, Shimamoto T. Directional difference of conduction velocity in the cardiac ventricular syncytium studied by microelectrodes. Circ Res. 1959;7:262-267.[Abstract/Free Full Text]

44. Myerburg RJ, Gelband H, Nilsson K, Castellanos A, Morales AR, Bassett AL. The role of canine superficial ventricular muscle fibers in endocardial impulse distribution. Circ Res. 1978;42:27-35.[Abstract/Free Full Text]

45. Weidmann S. The effect of the cardiac membrane potential on the rapid availability of the sodium-carrying system. J Physiol (Lond). 1955;127:213-224.

46. Frazier DW, Wolf PD, Wharton JM, Tang ASL, Smith WM, Ideker RE. Stimulus-induced critical point: mechanism for electrical initiation of reentry in normal canine myocardium. J Clin Invest. 1989;83:1039-1052.

47. Winfree AT. When time breaks down: the three-dimensional dynamics of electrochemical waves and cardiac arrhythmias. Princeton, NJ: Princeton University Press; 1987:125-153.

48. Moe GK, Rheinboldt WC, Abildskov JA. A computer model of atrial fibrillation. Am Heart J. 1964;67:200-220.[Medline] [Order article via Infotrieve]

49. Wiggers CJ. The mechanism and nature of ventricular fibrillation. Am Heart J. 1940;20:399-412.

50. Fabritz CL, Kirchhof PF, Fletcher RD, Zabel M, Franz MR. Postshock repolarization time dispersion measured by two monophasic action potential recordings predicts defibrillation success in humans. Circulation. 1994;90(pt 2):I-446. Abstract.

51. Shibata N, Chen PS, Dixon EG, Wolf PD, Danieley ND, Smith WM, Ideker RE. Epicardial activation after unsuccessful defibrillation in dogs. Am J Physiol. 1988;255:H902-H909.[Abstract/Free Full Text]

52. Bernstein RC, Frame LH. Ventricular reentry around a fixed barrier: resetting with advancement in an in vitro model. Circulation. 1990;81:267-280.[Abstract/Free Full Text]

53. Boersma L, Brugada J, Kirchhof C, Allesie M. Mapping of reset of anatomic and functional reentry in anisotropic rabbit ventricular myocardium. Circulation. 1994;89:852-862.[Abstract/Free Full Text]

54. Cha Y-M, Birgersdotter-Green U, Wolf PL, Peters BB, Chen P-S. The mechanism of termination of reentrant activity in ventricular fibrillation. Circ Res. 1994;74:495-506.[Abstract/Free Full Text]

55. Kirchhof C, Chorro F, Scheffer GJ, Brugada J, Konings K, Zetelaki Z, Allessie M. Regional entrainment of atrial fibrillation studied by high-resolution mapping in open-chest dogs. Circulation. 1993;88:736-749.[Abstract/Free Full Text]

56. KenKnight BH, Bayly PV, Gerstle RJ, Rollins DL, Wolf PD, Smith WM, Ideker RE. Regional capture of fibrillating ventricular myocardium: evidence of an excitable gap. Circ Res. 1995;77:849-855.[Abstract/Free Full Text]

57. Ebihara L, Johnson EA. Fast sodium current in cardiac muscle: a quantitative description. Biophys J. 1980;32:779-790.[Medline] [Order article via Infotrieve]

58. Hoffman BF, Kao CY, Suckling EE. Refractoriness in cardiac muscle. Am J Physiol. 1957;190:473-482.

59. Sweeney RJ, Gill RM, Steinberg MI, Reid PR. Ventricular refractory period extension caused by defibrillation shocks. Circulation. 1990;82:965-972.[Abstract/Free Full Text]

60. Dillon SM, Mehra R. Prolongation of ventricular refractoriness by defibrillation shocks may be due to additional depolarization of the action potential. J Cardiovasc Electrophysiol. 1992;3:442-456.

61. Swartz JF, Jones JL, Jones RE, Fletcher R. Conditioning prepulse of biphasic defibrillator waveforms enhances refractoriness to fibrillation wavefronts. Circ Res. 1991;68:438-449.[Abstract/Free Full Text]

62. Delmar M, Michaels DC, Johnson T, Jalife J. Effects of increasing intercellular resistance on transverse and longitudinal propagation in sheep epicardial muscle. Circ Res. 1987;60:780-785.[Abstract/Free Full Text]

63. Delgado C, Steinhaus B, Delmar M, Chialvo DR, Jalife J. Directional differences in excitability and margin of safety for propagation in sheep ventricular epicardial muscle. Circ Res. 1990;67:97-110.[Abstract/Free Full Text]

64. Spach MS, Miller WT, Geselowitz DB, Barr RC, Kootsey JM, Johnson EA. The discontinuous nature of propagation in normal canine cardiac muscle: evidence for recurrent discontinuities of intracellular resistance that affect the membrane currents. Circ Res. 1981;48:39-54.[Free Full Text]

65. Frazier DW, Krassowska W, Chen P-S, Wolf PD, Dixon EG, Smith WM, Ideker RE. Extracellular field required for excitation in three-dimensional anisotropic canine myocardium. Circ Res. 1988;63:147-164.[Abstract/Free Full Text]

66. Winfree AT. The electrical thresholds of ventricular myocardium. J Cardiovasc Electrophysiol. 1990;1:393-410.

67. 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.[Abstract/Free Full Text]

This article has been cited by other articles:

				J. P. Daubert and S.-S. Sheu Mystery of Biphasic Defibrillation Waveform Efficacy: Is it Calcium? J. Am. Coll. Cardiol., September 2, 2008; 52(10): 836 - 838. [Full Text] [PDF]

				Y.-H. Choi, C. Stamm, P. E. Hammer, K. F. Kwaku, J. J. Marler, I. Friehs, M. Jones, C. M. Rader, N. Roy, M.-T. Eddy, et al. Cardiac Conduction through Engineered Tissue Am. J. Pathol., July 1, 2006; 169(1): 72 - 85. [Abstract] [Full Text] [PDF]

				N. Trayanova Defibrillation of the heart: insights into mechanisms from modelling studies Exp Physiol, March 1, 2006; 91(2): 323 - 337. [Abstract] [Full Text] [PDF]

				A A J Adgey and S J Walsh Theory and practice of defibrillation: (1) Atrial fibrillation and DC conversion Heart, December 1, 2004; 90(12): 1493 - 1498. [Full Text] [PDF]

				N. Chattipakorn, I. Banville, R. A Gray, and R. E Ideker Effects of shock strengths on ventricular defibrillation failure Cardiovasc Res, January 1, 2004; 61(1): 39 - 44. [Abstract] [Full Text] [PDF]

				J. J. Sims, K. L. Schoff, J. M. Loeb, and N. A. Wiegert Regional gap junction inhibition increases defibrillation thresholds Am J Physiol Heart Circ Physiol, June 5, 2003; 285(1): H10 - H16. [Abstract] [Full Text] [PDF]

				M. Yashima, Y.-H. Kim, S. Armin, T.-J. Wu, Y. Miyauchi, W. J. Mandel, P.-S. Chen, and H. S. Karagueuzian On the mechanism of the probabilistic nature of ventricular defibrillation threshold Am J Physiol Heart Circ Physiol, January 1, 2003; 284(1): H249 - H255. [Abstract] [Full Text] [PDF]

				A. Hamzei, T. Ohara, Y.-H. Kim, M.-H. Lee, O. Voroshilovski, S.-F. Lin, J. N. Weiss, P.-S. Chen, and H. S. Karagueuzian The Role of Approximate Entropy in Predicting Ventricular Defibrillation Threshold Journal of Cardiovascular Pharmacology and Therapeutics, March 1, 2002; 7(1): 45 - 52. [Abstract] [PDF]

				N. Chattipakorn, I. Banville, R. A. Gray, and R. E. Ideker Mechanism of Ventricular Defibrillation for Near-Defibrillation Threshold Shocks: A Whole-Heart Optical Mapping Study in Swine Circulation, September 11, 2001; 104(11): 1313 - 1319. [Abstract] [Full Text] [PDF]

				F. G. Akar, B. J. Roth, and D. S. Rosenbaum Optical measurement of cell-to-cell coupling in intact heart using subthreshold electrical stimulation Am J Physiol Heart Circ Physiol, August 1, 2001; 281(2): H533 - H542. [Abstract] [Full Text] [PDF]

				N. C. Wang, M.-H. Lee, T. Ohara, Y. Okuyama, G. A. Fishbein, S.-F. Lin, H. S. Karagueuzian, and P.-S. Chen Optical Mapping of Ventricular Defibrillation in Isolated Swine Right Ventricles: Demonstration of a Postshock Isoelectric Window After Near-Threshold Defibrillation Shocks Circulation, July 10, 2001; 104(2): 227 - 233. [Abstract] [Full Text] [PDF]

				H. S Karagueuzian and P.-S. Chen Cellular mechanism of reentry induced by a strong electrical stimulus: Implications for fibrillation and defibrillation Cardiovasc Res, May 1, 2001; 50(2): 251 - 262. [Abstract] [Full Text] [PDF]

				A. Al-Khadra, V. Nikolski, and I. R. Efimov The Role of Electroporation in Defibrillation Circ. Res., October 27, 2000; 87(9): 797 - 804. [Abstract] [Full Text] [PDF]

				N. Chattipakorn, P. C. Fotuhi, X. Zheng, and R. E. Ideker Left Ventricular Apex Ablation Decreases the Upper Limit of Vulnerability Circulation, May 30, 2000; 101(21): 2458 - 2460. [Abstract] [Full Text] [PDF]

				L. C. Baker, B. London, B.-R. Choi, G. Koren, and G. Salama Enhanced Dispersion of Repolarization and Refractoriness in Transgenic Mouse Hearts Promotes Reentrant Ventricular Tachycardia Circ. Res., March 3, 2000; 86(4): 396 - 407. [Abstract] [Full Text] [PDF]

				M. R. Ujhelyi, J. J. Sims, and A. W. Miller Induction of Electrical Heterogeneity Impairs Ventricular Defibrillation : An Effect Specific to Regional Conduction Velocity Slowing Circulation, December 21, 1999; 100(25): 2534 - 2540. [Abstract] [Full Text] [PDF]

				Y. Cheng, K. A. Mowrey, D. R. Van Wagoner, P. J. Tchou, and I. R. Efimov Virtual Electrode-Induced Reexcitation : A Mechanism of Defibrillation Circ. Res., November 26, 1999; 85(11): 1056 - 1066. [Abstract] [Full Text] [PDF]

				J. J. Sims, A. W. Miller, and M. R. Ujhelyi Disparate effects of biphasic and monophasic shocks on postshock refractory period dispersion Am J Physiol Heart Circ Physiol, June 1, 1998; 274(6): H1943 - H1949. [Abstract] [Full Text] [PDF]

				N. Chattipakorn, B. H. KenKnight, J. M. Rogers, R. G. Walker, G. P. Walcott, D. L. Rollins, W. M. Smith, and R. E. Ideker Locally Propagated Activation Immediately After Internal Defibrillation Circulation, April 14, 1998; 97(14): 1401 - 1410. [Abstract] [Full Text] [PDF]

				M. R. Ujhelyi, J. J. Sims, and A. W. Miller High-dose lidocaine does not affect defibrillation efficacy: implications for defibrillation mechanisms Am J Physiol Heart Circ Physiol, April 1, 1998; 274(4): H1113 - H1120. [Abstract] [Full Text] [PDF]

				A. P. Winecoff, J. J. Sims, M. L. Markel, and M. R. Ujhelyi Pinacidil's Effects on Defibrillation Outcomes: Role of Increased Potassium Conductance Via the KATP Channel Journal of Cardiovascular Pharmacology and Therapeutics, January 1, 1997; 2(3): 171 - 180. [Abstract] [PDF]

				D. Vaidya, H. S. Tamaddon, C. W. Lo, S. M. Taffet, M. Delmar, G. E. Morley, and J. Jalife Null Mutation of Connexin43 Causes Slow Propagation of Ventricular Activation in the Late Stages of Mouse Embryonic Development Circ. Res., June 8, 2001; 88(11): 1196 - 1202. [Abstract] [Full Text] [PDF]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kwaku, K. F. Articles by Dillon, S. M. Search for Related Content PubMed PubMed Citation Articles by Kwaku, K. F. Articles by Dillon, S. M.