Original Articles

Genesis of Phase 3 Early Afterdepolarizations and Triggered Activity in Acquired Long-QT SyndromeClinical Perspective

Mitsunori Maruyama, Shien-Fong Lin, Yuanfang Xie, Su-Kiat Chua, Boyoung Joung, Seongwook Han, Tetsuji Shinohara, Mark J. Shen, Zhilin Qu, James N. Weiss, Peng-Sheng Chen
https://doi.org/10.1161/CIRCEP.110.959064
Circulation: Arrhythmia and Electrophysiology. 2011;4:103-111
Originally published February 15, 2011

Jump to

Abstract

Background— Both phase 2 and phase 3 early afterdepolarizations (EADs) occur in long-QT syndromes, but their respective roles in generating arrhythmias in intact cardiac tissue are incompletely understood.

Methods and Results— Intracellular Ca (Cai) and membrane voltage (Vm) were optically mapped in a quasi 2-dimensional model of cryoablated Langendorff-perfused rabbit ventricles (n=16). E-4031 (an IKr blocker) combined with reduced extracellular K ([K+]o) and Mg ([Mg2+]o) prolonged action potential duration heterogeneously and induced phase 2 and phase 3 EADs. Whereas phase 2 EADs were Cai-dependent, phase 3 EADs were not. The origins of 47 triggered activity episodes were attributed to phase 2 EADs in 12 episodes (26%) and phase 3 EADs in 35 episodes (74%). When phase 2 EADs accompanied phase 3 EADs, they accentuated action potential duration heterogeneity, creating a large Vm gradient across the boundary between long and short action potential duration regions from which triggered activity emerged. The amplitude of phase 3 EADs correlated with the Vm gradient (r=0.898, P<0.001). Computer simulation studies showed that coupling of cells with heterogeneous repolarization could extrinsically generate phase 3 EADs via electrotonic current flow. Alternatively, reduced IK1 caused by low [K+]o could generate intrinsic phase 3 EADs capable of inducing triggered activity at the boundary zone.

Conclusions— Phase 3 EADs can be extrinsic as the result of electrotonic current across steep repolarization gradients or intrinsic as the result of low IK1 and do not require spontaneous sarcoplasmic reticulum Ca release. Reduction of IK1 by low [K+]o strongly promotes ventricular arrhythmias mediated by phase 3 EADs in acquired long-QT syndrome caused by IKr blockade.

Introduction

Early afterdepolarizations (EADs) are strongly associated with ventricular arrhythmias in long-QT syndromes (LQTS). Both phase 2 and phase 3 EADs have been described in LQTS. It is generally accepted that phase 2 EADs result from the reactivation of ICa,L and/or spontaneous Ca release from the sarcoplasmic reticulum (SR).1–3 The ionic mechanism of phase 3 EADs, however, is less clear. Whereas phase 2 EADs can be readily recorded from isolated myocytes as well as intact tissue, phase 3 EADs have usually been reported in intact tissue preparations, such as Purkinje fibers or ventricular muscle.4–8 It has been suggested that spontaneous SR Ca release may underlie phase 3 EADs because they may occur concurrently with delayed afterdepolarizations,4–5 are facilitated by intracellular Ca (Cai) loading, and are suppressed by inhibition of Na+/Ca2+exchanger current (INCX).1,5 Alternatively, it is possible that most phase 3 EADs observed in tissue are not genuine cellular-level phenomena but are a consequence of “prolonged repolarization-dependent reexcitation.” Brugada and Wellens9 conjectured that if dispersion of repolarization is enhanced in LQTS, a voltage gradient between long and short action potential duration (APD) regions could create a “boundary” current that electrotonically depolarizes the short APD region as it tries to repolarize, generating triggered activity (TA) as long as a large voltage gradient is maintained. This form of TA arising from the boundary zone was observed in a partition chamber that artificially created heterogeneous repolarization,10 but it has not been demonstrated in the setting of LQTS in intact hearts. We tested this hypothesis by performing high-resolution optical mapping of Cai and membrane voltage (Vm) in a rabbit model of acquired LQTS, accompanied by computer simulations that reproduced the experimental observations. Our combined experimental and modeling findings suggest that electrotonic interactions between a long APD region, with or without phase 2 EADs, and its neighboring repolarizing region, is a major cause of phase 3 EADs and TA in this model. There was no evidence that phase 3 EADs were dependent on changes in Cai, suggesting that they are more likely to result from electrotonic interactions across boundaries with steep APD gradients rather than local spontaneous SR Ca release.

Clinical Perspective on p 111

Methods

Two-Dimensional Epicardial Layer of Langendorff-Perfused Rabbit Ventricles

This study protocol was approved by the Institutional Animal Care and Use Committee of Indiana University School of Medicine and conforms to the guidelines of the American Heart Association. New Zealand White female adult rabbits (n=16) were anesthetized with sodium pentobarbital (50 mg/kg). The heart was rapidly excised and perfused at 25 to 30 mL/min using a Langendorff perfusion system with oxygenated Tyrode solution (in mmol/L: NaCl 125, KCl 4.5, NaHCO3 24, NaH2PO4 1.8, CaCl2 1.8, MgCl2 0.5, and glucose 5.5) with a pH of 7.40±0.05. We performed cryoablation using a 7-cm SurgiFrost probe (CryoCath Technologies Inc, Quebec, Canada) with the probe temperature of −135°C for 3 minutes for the right ventricle (RV) and 5 minutes for the left ventricle (LV). During the cryoablation, the epicardium was protected by immersing the heart into warm (37°C) Tyrode solution. After the study, we sectioned the heart horizontally to confirm the surviving epicardial layer (≈1 mm) with 1% triphenyltetrazolium chloride (Online Supplement Figure I).

Experiment Protocol

Simultaneous Cai and Vm mapping was performed as described previously.11 EADs and TAs were induced by bradycardia after endocardial cryoablation, a selective IKr blocker E-4031 (0.5 μmol/L, Tocris Bioscience, Ellisville, MO), and 50% reduction of extracellular potassium ([K+]o) and magnesium ([Mg2+]o) concentrations.3 We also examined the effects of BAPTA-AM (20 μmol/L, Tocris) to determine the role of Cai in EADs (n=4). A pseudo-ECG was recorded by using the chamber solution as a volume conductor. Bipolar electrodes were attached to the RV outflow tract for pacing. In 3 hearts, transmembrane potentials were recorded during optical mapping with a standard glass microelectrode, as described previously.11 The heart was immersed in the heated tissue chamber (37°C) throughout the study to avoid a temperature gradient of the mapped surface. Contraction was inhibited with 10 μmol/L blebbistatin (Tocris) during optical mapping.

Definitions

Cai Reelevation

Cai reelevation was defined as an increase in Cai during the plateau of action potential (AP) after the first upstroke of Cai transient.

Phase 2 EAD

Phase 2 EAD was defined as depolarizing afterpotentials occurring at the plateau level of AP. Depolarization during the plateau phase was not considered as phase 2 EAD if the depolarization was induced by electrotonic interaction of an activation wave fronts passing near the pixel of interest.

Phase 3 EAD

Phase 3 EAD was defined as afterpotentials that retard the expected course of phase 3 AP repolarization. Previous studies consistently defined phase 3 EADs by this criterion, even though no actual depolarization is observed.4–8

Triggered Activity

TA was defined as propagated responses evoked by EADs that inscribe a QRS complex on the pseudo-ECG.

Torsade de Pointes

Torsade de pointes (TdP) was defined as polymorphic ventricular tachycardia (VT) characterized by undulating changes in the QRS axis on the pseudo-ECG.

Data Analysis

The methods for optical data processing were reported elsewhere.11 We defined the amplitude of baseline Vm and Cai transient as 1 arbitrary unit (AU). The Vm phase map was made using a time-delay embedding method.12 To construct Vm gradient (VG) map, the VG at the pixel with coordinates (n, m) was calculated with the larger of the absolute value of [Vm(n−3, m)−Vm(n+3, m)/distance and Vm(n, m−3)−Vm(n, m+3)]/distance. Because the interpixel distance was 0.35 mm, the pixel distance used in the gradient calculation more than 6 pixels was 2.1 mm. APD was measured at 70% repolarization. APD dispersion was defined as a time difference between maximal and minimal APDs in the mapped area. The amplitude of phase 3 EAD was measured as the difference between the diastolic resting Vm and the first deviation from smooth contour during phase 3 repolarization.6

Continuous variables are expressed as mean±SEM unless otherwise indicated. Statistical analysis was performed by paired t test to compare VG values with and without TAs and by 1-way repeated-measures ANOVA to compare the data for different pacing cycle lengths. The relationship between maximal VG and APD dispersion and between phase 3 EAD amplitude and VG were tested with the Pearson correlation. P≤0.05 was considered statistically significant.

Computer Simulation

Simulations corresponding to isolated myocytes and 1-dimensional (1D) cables were performed. AP models in the simulations were modified from the rabbit ventricular AP model developed by our group.13,14 To model the effects of E-4031 and hypokalemia, IKr was blocked and the extracellular potassium concentration [K+]0 was reduced to 2.7 mmol/L. The details of the AP models are shown in Online Supplemental Materials. For the isolated myocytes, the governing differential equation for voltage is CmdVdt=−(Iion+Istim) (1) where Cm is the membrane capacitance set at 1 μF/cm2, Iion is the total membrane ionic currents, and Istim is the stimulation current, with a square pulse of strength 40 μA/cm2and duration of 1 ms. The differential equation for voltage in the 1D cable is ∂V∂t=−Iion/Cm+D∂2V∂x2 (2) where D is the diffusion coefficient with the value of 0.0002 cm2/ms. No-flux boundary conditions were used. Numerically, Equation 2 was discretized with Δx=Δy=0.015 cm and integrated with a forward Euler method with an adaptive time step varying from 0.01 ms to 0.1 ms.

Results

Phase 2 EADs and Heterogeneous Repolarization

Exposure to 0.5 μmol/L E-4031 in combination with a 50% reduction in [K+]o and [Mg2+]o prolonged QT intervals and induced R-on-T ectopic beats in all hearts studied (n=16). In 6 of 16 hearts, data were also collected before and shortly (approximately 5 minutes) after lowering [K+]o and [Mg2+]o to measure QT interval and APD during stable pacing. R-on-T ectopic beats usually appeared 5 to 15 minutes after reducing [K+]o and [Mg2+]o. Because these ectopic beats caused irregular cycle lengths, QT and APD were not measured when there were ectopic activities. Changes in QT interval and APD are summarized in the Table. As the QT interval was prolonged, both APD dispersion and maximal VG during repolarization increased. The maximal VG correlated with APD dispersion (Online Supplement Figure II). APD prolongation was spatially heterogeneous, causing “island-like” long APD regions to emerge (Figure 1A), with a large VG at the boundary zone between the long and short APD regions. The size of the long APD region was dependent on cycle length but independent of activation sequence (Online Supplement Figure IIIA). The spatial distribution of long APD regions varied among individual hearts, although gradual changes in the distribution occurred over time in the same heart (Online Supplement Figure IIIB). Phase 2 EADs and R-on-T ventricular ectopic beats were observed after reducing [K+]o and [Mg2+]o. Phase 2 EADs further enhanced heterogeneity of repolarization because the phase 2 EADs occurred exclusively in long APD regions.

View this table:
Table.

Effects of E-4031 and Low [K+]o and [Mg2+]o on QT Intervals, APD, APD Dispersion, and VG (n=6)

Figure 1.
Figure 1.

Development of phase 2 and phase 3 EADs. A, The spatial distribution of APDs and VGs during repolarization is displayed in color-scaled maps. Optical APs at the maximal (filled squares) and minimal (unfilled squares) APD sites are shown with pseudo-ECG (pECG). In addition to E-4031, reducing [K+]o and [Mg2+]o greatly enhanced repolarization heterogeneity. Note that the appearance of phase 2 EAD (filled circle) further increased the VGs and was associated with emergence of an R-on-T ventricular ectopic beat (asterisk). APD was not measurable in a blank area of the APD map, where repolarization was interrupted by the ectopic beat. B, With E-4031 alone, there was no phase 3 EAD. Phase 3 EAD (unfilled circles) and an ectopic beats (asterisk) emerged after reduction in [K+]o and [Mg2+]o. Phase 3 EAD is discernable as the Vm difference between the resting Vm and the first deviation from the smooth contour during phase 3 repolarization.

Phase 3 EADs

Phase 3 EADs became manifest after E-4031 infusion with low [K+]o and [Mg2+]o (Figure 1B). The cycle length dependence of phase 3 EADs was comparable to previously reported observations (Online Supplement Figure IV).8,15 Phase 3 EADs were also confirmed in transmembrane potential recordings (Online Supplement Figure V). No transition from phase 2 EADs to phase 3 EADs was observed (Online Supplement Figure VI). Simultaneous recording of Vm and Cai during phase 2 and phase 3 EADs (Figure 2A) revealed that the onset of phase 2 EADs did not precede that of Cai reelevation at the site of phase 2 EAD origin, as reported previously.3 Of 55 episodes of phase 2 EAD arising from the mapped area, Cai reelevations occurred earlier than phase 2 EADs by 26±2 ms in 44 episodes (80%) or synchronously in 11 episodes (20%). In contrast, changes in Vm always preceded changes in Cai during phase 3 EADs by 25±3 ms, suggesting that spontaneous SR Ca release may underlie the development of phase 2 EADs but not for phase 3 EADs. To further confirm the role of Cai in EADs, we tested the effect of the Cai chelator BAPTA-AM (20 μmol/L, n=4). The maximal amplitude of Cai transient was decreased by 69±4% after 60 minutes of BAPTA-AM infusion. BAPTA-AM abolished phase 2 EADs, but phase 3 EADs persisted after BAPTA-AM loading in all hearts studied (Figure 2B). These findings suggest that whereas phase 2 EADs are Cai-dependent, phase 3 EADs are not. Interestingly, the largest phase 3 EADs always occurred at the boundary between long and short APD regions (Figure 3A). At the site with the largest phase 3 EAD, the EAD amplitude correlated with the VG at the time of the EAD onset (r=0.898, P<0.001, Figure 3B). Taken together, the findings suggest that electrotonic currents flowing from more positive Vm in long APD regions to shorter APD regions can cause phase 3 EADs at the boundary zone, without any requirement for SR Ca release.

Figure 2.
Figure 2.

Role of Cai in EADs. A, Simultaneous Cai and Vm recordings for phase 2 (black filled circle) and phase 3 (unfilled circle) EADs. Left panel, Onset of Cai reelevation (red arrow) preceded that of phase 2 EAD (black arrow). Right panel, First deviation of Vm (black arrow) was followed by the deviation of Cai (red arrow) from the expected trajectory. B, Effect of BAPTA-AM (20 μmol/L) on EADs. BAPTA-AM abolished phase 2 EAD (black filled circle) but not phase 3 EAD (unfilled circles).

Figure 3.
Figure 3.

Phase 3 EAD and VG. A, Spatial distribution of phase 3 EADs. Vm tracings were obtained at sites indicated in the APD map. Blue vertical line denotes the timing for the onset of the largest EAD at the boundary site (site 3). Note a high level of VG at site 3 at this timing. B, Relationship of the amplitude of phase 3 EADs and VG at the EAD onset.

EAD-Mediated TAs

No TAs occurred with E-4031 alone, but 114 spontaneous episodes of TAs were observed after a 50% reduction in [K+]o and [Mg2+]o. Forty-seven TA episodes (41%) originated from the mapped area, with 3 modes of initiation identified, as described below.

First, TAs could be directly induced by phase 2 EADs. Figure 4 shows an example. During the early phase of repolarization, Vm and Cai levels increased again (ie, phase 2 EAD and Cai reelevation, respectively) near the center of the long APD region (site a, squares in Vm/Cai ratio maps). The phase 2 EAD spread within the long APD region but did not directly propagate beyond. However, TA emerged from the boundary between long and short APD regions where excitability had recovered (site c, yellow arrowhead). TA rapidly propagated through the more repolarized area (ie, short APD region), which inscribed the QRS complex on the pseudo-ECG. The second hump (asterisks) was generated in long APD region as the TA wave front traveled around the long APD region, suggesting an electrotonic depolarization.

Figure 4.
Figure 4.

Phase 2 EAD–mediated TA. Left panels, Snapshots of Cai and Vm ratio maps at times indicated in the numbers on pseudo-ECG. Right panels, Cai and Vm tracings at sites indicated on the schematic diagram. Long and short APD regions are shaded and unshaded, respectively. Red vertical lines denote the QRS onset of the TA. Vm tracing at the earliest activation site of the TA (site c) is shown in blue. Note that propagated phase 2 EADs (black circles) were accompanied by Cai reelevations (red circles) and directly induced the TA at the boundary zone.

Second, TAs could arise directly from phase 3 EADs in the absence of phase 2 EADs. Figure 5 shows an example of TA mediated by a phase 3 EAD. APD in the mid-LV epicardium was heterogeneously prolonged by E-4031 with low [K+]o and [Mg2+]o. There were no phase 2 EADs in long APD region (site a). TA then emerged 366 ms after ventricular pacing at a pacing cycle length of 2000 ms (yellow arrowhead). Note that the earliest activation did not reside in long APD region (site a) but occurred at the inferior boundary between long and short APD regions (site b). Phase maps revealed that the course of repolarization (in yellow to red) at the inferior boundary suddenly reversed, generating a phase 3 EAD that induced TA. The VG across the boundary was the maximal at that time.

Figure 5.
Figure 5.

Phase 3 EAD-mediated TA. A, Snapshots of Vm ratio maps at times after pacing. B, Phase maps in the same episode. The area where the TA occurred is magnified. C, Vm tracings recorded at sites indicated in the phase map. Blue line shows Vm tracing at the earliest activation site of the TA (site b). S indicates pacing stimulus. D, Superimposed Vm tracings of long APD region (site a) and the TA origin (site b). VG map just before the TA initiation is also shown. Note a large VG and the presence of phase 3 EADs (unfilled circles) at site b.

Third, phase 3 EAD–mediated TAs could also occur in association with phase 2 EADs. In Figure 6, a phase 2 EAD and Cai reelevation occurred (squares in Vm/Cai ratio maps) from the long APD region located at the basal LV and RV. Afterward, TA emerged from the boundary zone (yellow arrowhead). However, phase 2 EADs (black filled circle) failed to cause TA, and a phase 3 EAD (black unfilled circle) was present at the earliest activation site of the TA (site b). The phase 3 EAD also coincided with a high VG, but in this case, phase 2 EAD aggravated the VG by further delaying repolarization in the long APD region (site a). Therefore, interaction between phase 2 and phase 3 EADs was important for this mode of TA initiation.

Figure 6.
Figure 6.

Interaction between phase 2 and phase 3 EADs. Left panels, Snapshots of Cai and Vm ratio maps at times indicated in the numbers on pseudo-ECG, phase maps showing the TA initiation (times denote intervals after the onset of escape beat, esc), and VG map just before the TA initiation. Right panels, Vm and Cai tracings recorded at sites indicated in the phase map. Note that phase 2 EAD (filled circle) was only in long APD region (site a). At the earliest site of the TA (site b, blue line), phase 3 EAD (unfilled circle) was present.

Of 47 episodes of TA, phase 2 EADs directly induced TA in 12 episodes (26%), whereas phase 3 EADs induced TA in 35 episodes (74%). The majority (63%) of the latter phase 3 EAD-mediated TA episodes were associated with electrotonic interaction with phase 2 EADs. The VG at the origin of the phase 3 EAD–mediated TA was the maximal or submaximal within the mapped area (average, 94±1%; range, 71% to 100% of the largest VG during repolarization). In 10 hearts in which the site of the TA origin was mapped and the maximal VG without TAs was measured during E-4031 infusion with reduced [K+]o and [Mg2+]o, the VG at the TA origin was significantly higher than the maximal VG in the absence of TA (0.40±0.03 AU/mm versus 0.34±0.03 AU/mm, P=0.016).

Mechanisms of VT

Thirty-eight episodes of VT were observed, including 8 monomorphic VTs, 25 polymorphic VTs, and 5 VTs with TdP-like morphological features. The mechanism of VT was determined in 18 episodes of VT originating from the mapped area (2 monomorphic VTs, 13 polymorphic VTs, and 3 TdPs). Repetitive phase 3 EAD–mediated TA arising from the same site at the boundary was responsible for all monomorphic VTs (Figure 7A, Movie I). Note that the long APD region never fully repolarized during VT (asterisks), probably because of electrotonic depolarization by TA (blue arrows), whereby the high VG was maintained. Two mechanisms accounted for polymorphic VT. Repetitive focal activations from single or multiple foci shifting from beat to beat caused 11 episodes (Figure 7B, Movie II). In the remaining 2 episodes, a combination of focal activations and macroreentry revolving around long APD region was responsible (Figure 7C, Movie III). Three episodes of TdP were induced by phase 3 EAD–mediated TA associated with phase 2 EADs (Online Supplement Figure VII), which initiated a reentrant rotor with a drifting core. We did not observe TdP maintained by a focal activation mechanism.

Figure 7.
Figure 7.

Mechanisms of VT. A, Monomorphic VT initiated and maintained by phase 3 EAD–mediated TAs. Vm tracings were recorded at long APD region (site a), the origin of the TAs at the boundary zone (site b), and short APD region (site c). Long APD regions and activation foci are illustrated in schematic diagram as shaded area and circles with arrows, respectively. Note that prominent phase 3 EADs (unfilled circles) were seen only at the boundary zone. Electrotonic transmission of more positive Vm at site a to site b (antegrade electrotonic interaction, red arrows) may cause the phase 3 EADs, whereas the resulting TAs may electrotonically depolarize site a (asterisks) by retrograde electrotonic interaction (blue arrows). esc indicates escape beat. B, Polymorphic VT with varying activation foci. Vm tracings were obtained at long (site d) and shorter (site e) APD regions. Phase 2 EADs (filled circles) contributed in part to sustenance of long APD region but was not essential for VT maintenance. C, Polymorphic VT with a combined mechanism of focal activation and macroreentry.

Computer Simulations

To explore the role of electrotonic coupling in the genesis of phase 3 EADs and TAs in tissue with heterogeneous APD caused by reduced repolarization reserve, we modified a rabbit ventricular AP model to exhibit 2 (for simplicity) types of AP morphology (Figure 8). When we coupled cells with short AP and prolonged AP with small phase 2 EADs, phase 3 EADs emerged at the boundary zone between these 2 types of cells in a 1D cable, resembling our experimental results as well as those in previous reports.4–8 Even with very large differences in APD, TA did not occur if [K+]o was normal. When [K+]o was reduced from 5.4 mmol/L to 2.7 mmol/L, delayed repolarization mimicking phase 3 EADs appeared due to the resulting reduction in IK1. The mechanism of low [K+]o inducing phase 3 EADs was detailed in a simulation study by Luo and Rudy.16 Reduction of IKr further potentiated the formation of phase 3 EADs as the result of its effect on reducing repolarization reserve. However, these phase 3 EADs also did not cause TA in uncoupled cells. When the cells were coupled under low [K+]o conditions, TA arose from the short APD side of the boundary zone despite a smaller APD difference between 2 types of cells. In line with the optical mapping results, no spontaneous Ca release was present during phase 3 EADs in the AP models.

Figure 8.
Figure 8.

Computer simulations. The mechanisms generating phase 3 EADs in a 1D cable of cardiac myocytes are shown. The 1D cable (200 cells=1.5 cm) contained 2 types of cells with different AP types when uncoupled (left traces, black and red, respectively). The cable was paced for 1 beat from site 1 to induce EADs and TA. A, Short APs (black) and long APs with multiple small phase 2 EADs (red) in a normal [K+]o condition. B, APs with short (black) and long (red) AP in a low [K+]o condition. Either electrotonic current across steep repolarization gradient or low [K+]o caused depolarizing hump consistent with phase 3 EADs (unfilled circles). Note that both effects were necessary for the development of TAs (asterisks) originating from the short APD side of the boundary zone (site 2).

Discussion

Phase 2 EADs and TA

The ability of an EAD to propagate is favored by a more negative EAD take-off potential, allowing greater recovery of the ICa,L to facilitate propagation. This is supported by the experimental observations of Damiano and Rosen,8 who reported that phase 3 EADs but not phase 2 EADs induced TA. On the other hand, Yan et al17 demonstrated that phase 2 EADs were directly associated with TA and TdP in a canine LV wedge preparation. However, the main vector of TA activation in their study was not from the phase 2 EAD site, raising uncertainty as to how phase 2 EADs elicited propagated responses. We found that when phase 2 EADs were confined to the center of the long APD region, they did not induce TA. When phase 2 EADs occurred near or were transmitted to the boundary zone, such that they encountered partially repolarized tissue, however, a new AP could be triggered and propagate into the short APD region (Figure 4). Therefore, our results suggest that heterogeneous repolarization is necessary for phase 2 EADs to induce TA. This agrees with the finding that increased dispersion of repolarization facilitates the ability of phase 2 EADs to generate TA.17

Mechanism of Phase 3 EADs

Previous studies have suggested that elevated Cai causing enhanced INCX might underlie phase 3 EADs.1,5 However, we observed neither persistent Cai elevation nor spontaneous Cai elevation during the phase 3 EADs in the present study. It has been shown that phase 3 EADs can also occur as the result of INCX activation by a large Cai transient that outlasts the end of an AP.18,19 However, this type of “late” phase 3 EAD occurs when APD is shortened rather than prolonged. Although INCX inhibition may suppress phase 3 EADs that occur during APD prolongation,5 the same intervention also attenuates phase 2 EADs,20 which diminishes the VG between long and short APD regions. The latter effects may suppress the phase 3 EAD indirectly. Therefore, suppression of phase 3 EAD by INCX inhibition does not necessarily indicate that phase 3 EAD is purely INCX dependent.

The higher prevalence of phase 3 EADs at slower heart rates7 with low [K+]o8 is compatible with our findings because these interventions augment the VG. We found that a large VG related to heterogeneous repolarization is essential for phase 3 EADs. Liu and Laurita21 reported that the breakthrough site of TA always occurred where the local repolarization gradient was the largest in the canine wedge model of LQTS. In the present study, we can exclude breakthrough activation from the deeper layer because the cells beyond ≈1 mm from the epicardial surface were cryoablated. Thus, TA could only arise from the epicardial site where a phase switch from repolarization to depolarization first occurred. A large VG and phase 3 EAD at the “arrhythmogenic” boundary suggest electrotonic reexcitation as the most likely underlying mechanism of TA (antegrade electrotonic interaction, red arrows in Figure 7). The TA caused by electrotonically assisted phase 3 EAD then rapidly propagates over the short APD region, which in turn extends APD electrotonically in the long APD region (retrograde electrotonic interaction, blue arrows in Figure 7). As a result of this ping-pong interaction, the long APD region is precluded from full repolarization even at faster heart rates, and the persistent high VG generates repetitive firing of phase 3 EAD–mediated TAs. This mechanism is consistent with “prolonged repolarization-dependent reexcitation,”9 in which bidirectional electrotonic interaction between long and short APD regions contributes to the maintenance of VT.

Phase 3 EADs have been reported mainly in Purkinje fibers and in the in vivo hearts with intracellular microelectrode or monophasic AP recordings.4–8,15,22 Although Cai-dependent phase 3 EADs have been reported in isolated cardiomyocytes under some experimental conditions (eg, K+-free Tyrode solution23), the fact that the great majority of phase 3 EADs have been recorded from multicellular preparations suggests that electrotonic interactions may be important for the development of phase 3 EAD in many long QT settings. If true, this implies that a large portion of phase 3 EADs described in the literature do not arise solely from the intrinsic ionic currents of the myocyte but represent a combination of electrotonic currents interacting with intrinsic ionic currents. Using computer simulation studies, we documented this scenario by showing that a phase 3 EAD can be produced by electrotonic coupling in tissue with a large APD gradient. In addition, inhibition of IK1 can be the intrinsic source of phase 3 EADs, but the synergistic effect of the intrinsic and extrinsic mechanisms of phase 3 depolarization may be essential for ventricular arrhythmogenesis in acquired LQTS.

Clinical Implications

Our results provide new insights into the mechanisms of arrhythmogenesis in drug-induced LQTS during hypokalemia and hypomagnesemia. Increasing heart rate with temporary pacing is a reasonable treatment for acquired LQTS because the repolarization heterogeneity is cycle length–dependent. Treatments targeting to phase 2 EADs may also suppress phase 3 EADs, which are major causes of TA.

Study Limitations

We used E-4031 with low [K+]o and [Mg2+]o as a model of acquired LQTS. Our findings may not be applicable to other types of LQTS. We investigated only the epicardium in this study. However, previous studies have shown that EADs originate preferentially from Purkinje cells15,22 and M cells.24 It is possible that “prolonged repolarization-dependent reexcitation” may contribute to the generation of phase 3 EADs and TAs in the Purkinje-muscle junction and the boundary between M cells and other tissues. Our computer simulations were carried out only in 1D cable. Simulating reentry-like behavior will require 2D or 3D tissue models. However, 1D simulation allows us to focus on the mechanism of EADs and TAs, which are the primary purpose of this study.

Conclusions

A number of studies have documented the importance of heterogeneous repolarization in the ventricular arrhythmogenesis in LQTS.1–3,9,14,18,22 The present study demonstrates that heterogeneous repolarization is indispensable not only for creating a functional substrate promoting reentry but also directly participates in generating the triggers that emerge from phase 2 and phase 3 EADs. In addition, we provide the first direct experimental evidence for “prolonged repolarization-dependent reexcitation” in intact cardiac muscle tissue.

Sources of Funding

This study was supported in part by National Institutes of Health grants P01 HL78931, R01 HL78932, and 71140; a Nihon Kohden/St Jude Medical electrophysiology fellowship (Dr Maruyama); a Postdoctoral Fellowship Award from American Heart Association, Western States Affiliate (Dr Xie); The Heart Rhythm Society Fellowship in Cardiac Pacing and Electrophysiology (Dr Shen); the Kawata and Laubisch Endowments (Dr Weiss); an American Heart Association Established Investigator Award (Dr Lin); and a Medtronic-Zipes endowment (Dr Chen).

Disclosures

CryoCath Technologies, Inc, provided the SurgiFrost probe.

Acknowledgments

We thank Erica Foster for secretarial assistance.

Footnotes

  • Received June 9, 2010.
  • Accepted November 3, 2010.

References

  1. 1.↵
    1. Volders PG,
    2. Vos MA,
    3. Szabo B,
    4. Sipido KR,
    5. de Groot SH,
    6. Gorgels AP,
    7. Wellens HJ,
    8. Lazzara R
    . Progress in the understanding of cardiac early afterdepolarizations and torsades de pointes: time to revise current concepts. Cardiovasc Res. 2000; 46:376–392.
    OpenUrlFREE Full Text
  2. 2.↵
    1. January CT,
    2. Chau V,
    3. Makielski JC
    . Triggered activity in the heart: cellular mechanisms of early after-depolarizations. Eur Heart J. 1991; 12 (Suppl F):4–9.
    OpenUrlAbstract/FREE Full Text
  3. 3.↵
    1. Choi BR,
    2. Burton F,
    3. Salama G
    . Cytosolic Ca2+triggers early afterdepolarizations and Torsade de Pointes in rabbit hearts with type 2 long QT syndrome. J Physiol. 2002; 543:615–631.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Patterson E,
    2. Szabo B,
    3. Scherlag BJ,
    4. Lazzara R
    . Early and delayed afterdepolarizations associated with cesium chloride-induced arrhythmias in the dog. J Cardiovasc Pharmacol. 1990; 15:323–331.
    OpenUrlPubMed
  5. 5.↵
    1. Szabo B,
    2. Kovacs T,
    3. Lazzara R
    . Role of calcium loading in early afterdepolarizations generated by Cs+in canine and guinea pig Purkinje fibers. J Cardiovasc Electrophysiol. 1995; 6:796–812.
    OpenUrlPubMed
  6. 6.↵
    1. Bailie DS,
    2. Inoue H,
    3. Kaseda S,
    4. Ben-David J,
    5. Zipes DP
    . Magnesium suppression of early afterdepolarizations and ventricular tachyarrhythmias induced by cesium in dogs. Circulation. 1988; 77:1395–1402.
    OpenUrlAbstract/FREE Full Text
  7. 7.↵
    1. Davidenko JM,
    2. Cohen L,
    3. Goodrow R,
    4. Antzelevitch C
    . Quinidine-induced action potential prolongation, early afterdepolarizations, and triggered activity in canine Purkinje fibers: effects of stimulation rate, potassium, and magnesium. Circulation. 1989; 79:674–686.
    OpenUrlAbstract/FREE Full Text
  8. 8.↵
    1. Damiano BP,
    2. Rosen MR
    . Effects of pacing on triggered activity induced by early afterdepolarizations. Circulation. 1984; 69:1013–1025.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Brugada P,
    2. Wellens HJ
    . Early afterdepolarizations: role in conduction block, “prolonged repolarization-dependent reexcitation,” and tachyarrhythmias in the human heart. Pacing Clin Electrophysiol. 1985; 8:889–896.
    OpenUrlCrossRefPubMed
  10. 10.↵
    1. Kupersmith J,
    2. Li ZY,
    3. Maldonado C
    . Marked action potential prolongation as a source of injury current leading to border zone arrhythmogenesis. Am Heart J. 1994; 127:1543–1553.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Maruyama M,
    2. Joung B,
    3. Tang L,
    4. Shinohara T,
    5. On YK,
    6. Han S,
    7. Choi EK,
    8. Kim DH,
    9. Shen MJ,
    10. Weiss JN,
    11. Lin SF,
    12. Chen PS
    . Diastolic intracellular calcium-membrane voltage coupling gain and postshock arrhythmias: role of Purkinje fibers and triggered activity. Circ Res. 2010; 106:399–408.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Gray RA,
    2. Pertsov AM,
    3. Jalife J
    . Spatial and temporal organization during cardiac fibrillation. Nature. 1998; 392:75–78.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Mahajan A,
    2. Shiferaw Y,
    3. Sato D,
    4. Baher A,
    5. Olcese R,
    6. Xie LH,
    7. Yang MJ,
    8. Chen PS,
    9. Restrepo JG,
    10. Karma A,
    11. Garfinkel A,
    12. Qu Z,
    13. Weiss JN
    . A rabbit ventricular action potential model replicating cardiac dynamics at rapid heart rates. Biophys J. 2008; 94:392–410.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Sato D,
    2. Xie LH,
    3. Sovari AA,
    4. Tran DX,
    5. Morita N,
    6. Xie F,
    7. Karagueuzian H,
    8. Garfinkel A,
    9. Weiss JN,
    10. Qu Z
    . Synchronization of chaotic early afterdepolarizations in the genesis of cardiac arrhythmias. Proc Natl Acad Sci U S A. 2009; 106:2983–2988.
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    1. El-Sherif N,
    2. Zeiler RH,
    3. Craelius W,
    4. Gough WB,
    5. Henkin R
    . QTU prolongation and polymorphic ventricular tachyarrhythmias due to bradycardia-dependent early afterdepolarizations: afterdepolarizations and ventricular arrhythmias. Circ Res. 1988; 63:286–305.
    OpenUrlAbstract/FREE Full Text
  16. 16.↵
    1. Luo CH,
    2. Rudy Y
    . A dynamic model of the cardiac ventricular action potential, II: afterdepolarizations, triggered activity, and potentiation. Circ Res. 1994; 74:1097–1113.
    OpenUrlAbstract/FREE Full Text
  17. 17.↵
    1. Yan GX,
    2. Wu Y,
    3. Liu T,
    4. Wang J,
    5. Marinchak RA,
    6. Kowey PR
    . Phase 2 early afterdepolarization as a trigger of polymorphic ventricular tachycardia in acquired long-QT syndrome : direct evidence from intracellular recordings in the intact left ventricular wall. Circulation. 2001; 103:2851–2856.
    OpenUrlAbstract/FREE Full Text
  18. 18.↵
    1. Burashnikov A,
    2. Antzelevitch C
    . Late-phase 3 EAD: a unique mechanism contributing to initiation of atrial fibrillation. Pacing Clin Electrophysiol. 2006; 29:290–295.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Ogawa M,
    2. Morita N,
    3. Tang L,
    4. Karagueuzian HS,
    5. Weiss JN,
    6. Lin SF,
    7. Chen PS
    . Mechanisms of recurrent ventricular fibrillation in a rabbit model of pacing-induced heart failure. Heart Rhythm. 2009; 6:784–792.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Milberg P,
    2. Pott C,
    3. Fink M,
    4. Frommeyer G,
    5. Matsuda T,
    6. Baba A,
    7. Osada N,
    8. Breithardt G,
    9. Noble D,
    10. Eckardt L
    . Inhibition of the Na+/Ca2+exchanger suppresses torsades de pointes in an intact heart model of long QT syndrome-2 and long QT syndrome-3. Heart Rhythm. 2008; 5:1444–1452.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Liu J,
    2. Laurita KR
    . The mechanism of pause-induced torsade de pointes in long QT syndrome. J Cardiovasc Electrophysiol. 2005; 16:981–987.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Nattel S,
    2. Quantz MA
    . Pharmacological response of quinidine induced early afterdepolarisations in canine cardiac Purkinje fibres: insights into underlying ionic mechanisms. Cardiovasc Res. 1988; 22:808–817.
    OpenUrlAbstract/FREE Full Text
  23. 23.↵
    1. Miura M,
    2. Ishide N,
    3. Numaguchi H,
    4. Takishima T
    . Diversity of early afterdepolarizations in guinea pig myocytes: spatial characteristics of intracellular Ca2+concentration. Heart Vessels. 1995; 10:266–274.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Burashnikov A,
    2. Antzelevitch C
    . Acceleration-induced action potential prolongation and early afterdepolarizations. J Cardiovasc Electrophysiol. 1998; 9:934–948.
    OpenUrlPubMed

Clinical Perspective

Acquired long-QT syndrome caused by drug-induced IKr inhibition is an important cause of sudden cardiac death attributed to polymorphic ventricular tachycardia. Both phase 2 and phase 3 early afterdepolarizations (EADs) may be responsible for initiating arrhythmias, but the mechanisms of phase 3 EAD remain unclear. We performed simultaneous dual optical mapping of intracellular Ca2+ and membrane voltage in Langendorff-perfused rabbit ventricles. The EADs were induced by IKr blockade and reduced extracellular K+ and Mg2+ concentrations. More than two-thirds of triggered activities were induced by phase 3 EADs. Although some suggested that spontaneous Ca2+ release from the sarcoplasmic reticulum underlie phase 3 EADs, our results revealed that phase 2 EADs but not phase 3 EADs were dependent on intracellular Ca2+. Heterogeneous repolarization enhanced by IKr blockade and the electrolyte disturbances facilitate the development of phase 3 EADs through electrotonic interaction. This phenomenon is consistent with the concept of “prolonged repolarization-dependent reexcitation.” We also documented that heterogeneous repolarization was essential for phase 2 EADs to trigger propagating ventricular responses. Increasing heart rate with temporary pacing suppressed phase 3 EADs by reducing the repolarization heterogeneity. Computer simulation study revealed that reduced K+ concentration is important in promoting phase 3 EADs and triggered activities. Because phase 2 EADs are responsible for inducing phase 3 EADs through electrotonic interactions, it is possible that treatments targeting Ca2+-dependent phase 2 EADs may suppress Ca2+-independent phase 3 EADs. This study is clinically relevant because it may lead to mechanism-based therapies of phase 3 EADs, a major contributor to arrhythmogenesis in acquired long-QT syndrome.

View Abstract

This Issue

Jump to

Article Tools

Share this Article

  • Email
  • Genesis of Phase 3 Early Afterdepolarizations and Triggered Activity in Acquired Long-QT SyndromeClinical Perspective
    Mitsunori Maruyama, Shien-Fong Lin, Yuanfang Xie, Su-Kiat Chua, Boyoung Joung, Seongwook Han, Tetsuji Shinohara, Mark J. Shen, Zhilin Qu, James N. Weiss and Peng-Sheng Chen
    Circulation: Arrhythmia and Electrophysiology. 2011;4:103-111, originally published February 15, 2011
    https://doi.org/10.1161/CIRCEP.110.959064
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Related Articles

Cited By...

Subjects