Rotors anchored by refractory islands drive Torsades de Pointes in an experimental model of electrical storm

Abstract

Background:

Electrical storm (ES) is a life-threatening emergency in patients at high risk of ventricular tachycardia/fibrillation (VT/VF), but the pathophysiology and molecular basis are poorly understood.

Objective:

To explore the electrophysiological substrate for experimental ES.

Methods:

A model was created by inducing chronic complete atrioventricular-block in defibrillator-implanted rabbits, which recapitulates QT-prolongation, Torsades-des-Pointes (TdP) and VF-episodes.

Results:

Optical mapping revealed island-like regions with action potential duration (APD) prolongation in the left ventricle (LV), leading to increased spatial APD-dispersion, in rabbits with ES (defined as ≥3 VF-episodes/24-h). The maximum APD and its dispersion correlated with the total number of VF-episodes in-vivo. TdP was initiated by an ectopic beat that failed to enter the island and formed a reentrant wave, and perpetuated by rotors whose centers swirled in the periphery of the island. Epinephrine exacerbated the island by prolonging APD and enhancing APD-dispersion, which was less evident after late Na⁺-current (I_Na-L) blockade with 10 μM ranolazine. Non-sustained VT in a non-ES rabbit heart with homogeneous APD prolongation resulted from multiple foci with an electrocardiographic morphology different from TdP driven by drifting rotors in ES-rabbit hearts. The neuronal Na⁺-channel subunit Na_V1.8 was upregulated in ES-rabbit LV-tissues and expressed within myocardium corresponding to the island location in optically mapped ES-rabbit hearts. The Na_V1.8-blocker A-803467 (10 mg/kg, i.v.) attenuated QT-prolongation and suppressed epinephrine-evoked TdP.

Conclusion:

A tissue-island with enhanced refractoriness contributes to the generation of drifting rotors that underlies ES in this model. Na_V1.8-mediated I_Na-L merits further investigation as a contributor to the substrate for ES.

Graphical abstract

Introduction

Electrical storm (ES), characterized by repetitive episodes (generally defined as 3 or more within 24 hours) of ventricular tachycardia and/or fibrillation (VT/VF), is a life-threatening complication of implantable cardioverter-defibrillator (ICD) therapy. However, the pathophysiology and molecular basis of ES are poorly understood. We have created and initially characterized a model of ES by inducing chronic complete atrioventricular block (CAVB) in ICD-implanted rabbits, causing QT-interval prolongation, repetitive Torsades de Pointes (TdP) and frequent VF episodes.¹ Using this model, we showed that ES is associated with activation of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) and phosphorylation changes in L-type Ca²⁺-channels and ryanodine receptor type 2-channels,¹ contributing to the generation of afterdepolarization-related triggered activity.² The cardiac Na₊-channel subunit Na_V1.5 (encoded by SCN5A) is another important target of CaMKII, phosphorylation of which is linked to arrhythmogenic late Na⁺-current (I_Na-L),² but the role of this system in ES has not been evaluated.¹ In the present study, the electrophysiological mechanisms of experimental ES were explored using high-resolution optical mapping. We report our findings here: drifting rotors in association with I_Na-L-mediated localized action potential duration (APD) prolongation acted as a driver of TdP and neural Na⁺-channel Na_V1.8 upregulation, rather than CaMKII-mediated Na_V1.5 hyperphosphorylation, was implicated in enhanced I_Na-L.

Methods

All animal-handling protocols were approved by the Animal-Experimentation Ethics Committees of Nagasaki University and Nagoya University. Details are provided in Supplemental Material.

Experimental model of electrical storm

Female New Zealand white/Japanese white rabbits (2.5-3.5 kg) were used. Chemical atrioventricular node ablation and ICD (Medtronic) implantation were performed (Figure 1A), as previously described.¹ They had a slow escape rhythm and QT-interval prolongation/abnormal QTU-complexes and developed non-sustained VT (NSVT) and VF episodes (Figures 1B and 1C). In the present study, we evaluated 50 CAVB rabbits with ICDs: ES (defined as ≥3 VF episodes within 24 hours) developed in 39 rabbits (47±8 VF episodes/rabbit for 105±9 days) and the remaining 11 had NSVT or VF episodes but not ES (non-ES rabbits). Nineteen normal rabbits were used as controls (CTLs). Figure 1D shows the flow chart of electrophysiological and biochemical studies with the numbers of animals.

Figure 1. Rabbit model of electrical storm.

(A) Complete atrioventricular (AV) block creation and implantable cardioverter-defibrillator (ICD) implantation. (B) CAN-patch electrode ECGs and electrograms (EGMs: 2 leads at the epicardial right ventricle) (C) Examples of non-sustained VT (NSVT) and ventricular fibrillation (VF) episode. TdP, Torsades de Pointes. (D) The flow chart of the experiments. ES, electrical storm; IHC, immunohistochemistry.

Optical mapping

Fluorescent action potential signals from subepicardial myocardium of Langendorff-perfused hearts were recorded by a high-speed camera at 1,000 frames/s and the phase mapping method was applied to analyze excitation wave front-tail dynamics during VT/VF, as previously described.³ Phase singularity (PS), the organizing rotational center of spiral wave re-entry, was identified in each of 1,000 consecutive images. Using recently-developed PS detection algorithm to quantify the phase-variance values in a phase map and identify the position of PS as its peak,⁴ an accumulation map of phase-variance was created by integrating 2,000 consecutive maps.

Statistical analysis

Results are presented as median and interquartile range (IQR; at the 25^th and 75^th percentiles) or mean±SEM (standard error of the mean). Nonparametric tests and mixed effects model were used for statistical comparisons. Two-tailed P<0.05 indicated statistical significance.

Results

Characteristics of rabbits used for optical mapping

The characteristics of 15 ES and 2 non-ES rabbits used for optical mapping experiments are shown in Table SI. The ES rabbits had 37±6 VF episodes for 97±8 days.

Localized regions with prominent APD prolongation in an ES heart

Figure 2 illustrates representative APD maps in an ES, a non-ES (ES #3, Non-ES #2, respectively, in Table SI) and a CTL heart at baseline and in the presence of 0.1 μM epinephrine, followed by 10 μM ranolazine application. The map for the ES heart at baseline was heterogeneous, showing regions with remarkably prolonged APDs at the LV base (Figure 2A), in contrast to homogeneous maps for the non-ES (Figure 2B) and the CTL heart (Figure 2C).

Figure 2. APD maps.

Action potential duration (APD) maps at a pacing cycle length of 1,000 msec at baseline (A, B and C) and under the perfusate containing 0.1 μM epinephrine (D, E and F) followed by 10 μM ranolazine (G, H and I) in an electrical storm (ES), a non-ES and a control (CTL) rabbit heart. Values are APD₉₀ and APD₉₀-dispersion (APD₉₀-D) (msec). A dotted oval (A) indicates island-like regions with prominent APD prolongation.

Responses to epinephrine followed by ranolazine application

Epinephrine caused prominent APD prolongation, particularly at the base, in the ES heart. The changes exacerbated the island-like region, leading to strikingly increased APD dispersion (Figure 2D). In the non-ES and CTL hearts, APD was prolonged homogeneously, leading to much less its dispersion (Figure 2E and 2F, respectively) than the ES heart.

The island in the ES heart was less evident in response to I_Na-L blockade with 10 μM ranolazine (Figure 2G). Ranolazine shortened APD at the base and prolonged it at the apex. In the non-ES and the CTL heart treated with ranolazine, APD₉₀ (APD at 90% repolarization) was prolonged homogeneously, respectively (Figure 2H and 2I). The differential response of the ES heart versus the non-ES and CTL hearts can likely be explained by the combined I_Kr (the rapid component of delayed rectifier K⁺-current) and I_Na-L blocking actions of ranolazine. The agent has an IC₅₀ (half-maximal inhibitory concentration) to inhibit I_Kr and I_Na-L at 13 and 17 μM, respectively, in rabbit ventricular myocytes⁵ and 10 μM ranolazine causes APD prolongation at the base in normal rabbit hearts.⁶ The APD shortening at the base in the ES heart suggests a predominant effect of I_Na-L blockade.

Tissue-island with long APD in ES hearts

APD prolongation was heterogeneous in ES hearts. Figure 3A and 3B illustrate representative APD-dispersion and island maps, which were assessed by the difference from the minimum APD₉₀, in 3 ES, 2 non-ES and 1 CTL hearts. Based on our experience that APD₉₀-dispersion is <15 ms in normal rabbit hearts, tissue-islands were defined as regions where APD₉₀ values are >30 ms longer than minimum APD₉₀. Maximum APD₉₀-dispersion in the 3 ES hearts shown was much greater than in the non-ES and the CTL hearts. An island-like region is clearly present in the 3 ES hearts, but scarcely in the non-ES and CTL hearts. Figure 3C shows overall data for these variables. Median maximum APD₉₀ and APD₉₀-dispersion was 332 ms (IQR: 283 to 394 ms) and 69 ms (IQR: 55 to 96 ms), respectively, in ES hearts (n=14), values much greater than those in CTL hearts (n=4) (189 ms [IQR: 178 to 223 ms] for APD₉₀, P<0.001; 6 ms [IQR: 4 to 8 ms] for APD₉₀-dispersion, P<0.001). Median size of the island was 420 mm² (IQR: 110 to 738 mm²) in ES (n=14) and 0 mm² in all CTL hearts (n=4) (P<0.001).

Figure 3. APD-dispersion and island maps.

(A and B) Action potential duration (APD₉₀), APD-dispersion (APD₉₀-D) and island maps in 3 electrical storm (ES), 2 non-ES and 1 control (CTL) rabbit hearts. (C) Overall data on maximum APD₉₀, maximum APD₉₀-D and island size in ES (n=14) and CTL group (n=4). Statistical analysis could not be performed for Non-ES group (n=2). Nonparametric Kolmogorov-Smirnov test was used for single comparisons between groups. ****P<0.001 vs. CTL. (D and E) Maximum APD₉₀ and APD₉₀-D values plotted against the total number of ventricular fibrillation (VF) episodes and against ES event-days in 14 ES rabbits. Pearson’s correlation coefficient was used to measure the strength of linear associations.

In addition, there were linear associations between ex vivo APD findings and in vivo arrhythmia indices (total number of VF episodes; ES event-days) in 14 ES rabbits (Figure 3D and 3E). Maximum APD₉₀ and APD₉₀-dispersion correlated with the number of VF episodes (r=0.69, P<0.01 and r=0.61, P<0.05, respectively) and with ES event-days (r=0.76; P<0.005 and r=0.64; P<0.05, respectively). Moreover, premature ventricular contractions (PVCs) occurred in 14 ES hearts, most of which originated outside the mapped area. Some PVCs occurring from phase 2 of the preceding action potential originated at regions adjacent to the island (data not shown).

Rotors as a driver of TdP

Figure 4A shows the phase map analyses for TdP, optically mapped in ES #15 heart. Thirty-two out of 49 beats (65%) during TdP had rotor-type activation patterns, including a counterclockwise rotor, clockwise rotor, dual rotor and figure-of-8 rotor. The remaining 17 beats (35%) had breakthrough and one way-propagation for 14 and 3 beats, respectively. In total, 46 beats (94% of activations mapped) showed activation patterns that could be determined by mapping inside the recording area of the anterior epicardial surface. The beat-to-beat intervals of 49 beats were almost identical regardless of the type of activation pattern. The accumulation map of phase-variance to identify PS distribution revealed that the high phase-variance value area (shown in red) was located at the boundary of regions with APD prolongation. These observations suggest that drifting rotors in the periphery of the localized APD prolongation region are a driver of TdP in this model. It is likely that the 17 beats with breakthrough and one way-propagation patterns were wavefronts propagating form intramural sources.

Figure 4. Phase map analyses for TdP.

(A) Torsades de pointes (TdP) in electrical storm (ES) #15 heart. Each of 49 beats analyzed is labelled with C (clockwise rotor), CC (counterclockwise rotor), D (dual rotor), F (figure-of-8), B (breakthrough) or O (one way-propagation). White dotted circle in accumulation map of phase-variance indicates tissue-island with repolarization delay (left in B). (B and C) Phase maps for TdP initiation and last 4 beats before the self-termination. Black/white circle indicate phase singularity (PS) with clockwise/counterclockwise rotation. See text for details.

Mechanisms of TdP initiation and termination

Phase maps for TdP initiation are illustrated in Figure 4B. The tissue-island with delayed repolarization played a crucial role in generation and annihilation of TdP-driving rotors. An ectopic beat propagating from the lateral side failed to enter the refractory regions at the LV base (as shown in maps at frames of 236 and 240 ms) and created several PSs (PS1 to PS5) around the island (328 ms). PS2-3 and PS4-5 adjacent to each other at the LV middle to apex regions disappeared by mutual annihilation, whereas PS1 with clockwise rotation survived by meandering along the island following a curvilinear trajectory, and subsequently maintained the rotational activity. The ectopic beat that triggered TdP had a spiky QRS-complex, suggesting origination from Purkinje fibers.

Figure 4C shows phase maps for the last 4 beats before self-termination. A counterclockwise rotor drifting at the island edge formed the 4^th and 3^rd beat from the last (7036, 7044 and 7188 ms). Wave propagation from the upper middle side (7216 ms) produced rotational activity with PS shifting-up followed by PS jumping toward the base in association with a smaller excitable gap, which typified the last 2 beats, respectively (7336 and 7476 ms). The wavefront was then trapped by surrounding refractoriness, leading to PS annihilation without any offspring wavelet (7512 ms). Rotational activity thus disappeared in a cul-de-sac, resulting in TdP termination.

Another example is illustrated in Figure 5. ES #1 heart had a VF/defibrillation event before optical signal recording. Before VF re-initiation, an accelerated escape rhythm was optically mapped. The APD₉₀ map was heterogeneous. The APD₉₀-dispersion maps revealed an irregular-shaped tissue-island at the base. Figure 5B shows TdP occurring pause-dependently and phase maps for its termination. Seven PSs were present at frames of 68 and 665 ms, 5 of which located at the border of the island. Mutual annihilation of 2 paired PSs, PS shifting and jumping, and trapping of the spiral tip by surrounding refractory tissue were detected. The accumulation map of phase-variance showed high values distributed in a shape almost identical to the refractory tissue-island.

Figure 5. Severe arrhythmic events in an optically mapped electrical storm heart.

(A) Ventricular fibrillation (VF)/defibrillation events and the action potential duration (APD₉₀) and APD-dispersion (APD₉₀-D) maps for an accelerated junctional rhythm beat. (B) Torsades de pointes (TdP), 4 phase maps, action potential signals and accumulation map of phase variance. Black/white circle indicates phase singularity (PS) with clockwise/counterclockwise rotation; black dots, irregular-shaped tissue-island. See text for details.

NSVT in a non-ES heart

Figure SI show the isochrone maps for NSVT in Non-ES #2 heart with homogeneous APD prolongation (Figure 2E). The NSVT had relatively narrow QRS-complexes with irregular cycle lengths, a morphology different from TdP-like VT (shown in Figure 4A). The isochrone maps revealed multiple breakthrough activations with centrifugal propagation but did not show reentrant activity. Note that multiple foci formed membrane voltage oscillation at the center. The findings obtained from the non-ES versus ES hearts suggest discrete mechanisms of VTs with different electrophysiological substrates.

Lidocaine suppresses epinephrine-evoked TdP

Epinephrine challenge was effective to evoke TdP in vivo (Figure SII). To assess the involvement of I_Na-L in epinephrine-evoked TdP, we administered a single lidocaine 2 mg/kg injection in 4 rabbits (Figure SIII). Epinephrine prolonged the QT-interval and increased the numbers of PVCs, causing short VT-runs and TdP. Lidocaine completely suppressed short VT-runs and TdP by shortening the QT-interval.

I_Na-L in ES rabbit LV myocytes

Figure 6A shows representative 30 μM tetrodotoxin (TTX)-sensitive current recordings in LV myocytes from a CTL and an ES rabbit and the corresponding mean data. Cell capacitance was larger in ES myocytes (178±15 pF) than in CTLs (112±7 pF, P<0.005), indicating cell hypertrophy. Peak I_Na was similar between CTL and ES myocytes (108±13 and 98±19 pA/pF, respectively, P=0.34), I_Na-L density averaged 0.54±0.14 pA/pF in ES myocytes, versus 0.27±0.06 pA/pF in CTL myocytes (P=0.0785). Both cardiac Na⁺-channel Na_V1.5 currents and neuronal Na⁺-channel Na_V1.8 currents are TTX-resistant. Studies in transfected cells have shown that a high TTX concentration (30 μM) suppresses Na_V1.5 and hardly influences Na_V1.8 currents.^7,8 In this study 30 μM TTX-resistant currents were present in CTL and ES myocytes, with the current densities of the late component being significantly larger in ES (0.18±0.05) versus CTL myocytes (0.02±0.01, P<0.05, Figure 6B).

Figure 6. Na⁺-currents and CaMKII, Nav1.5 and Nav1.8 expression.

(A and B) Representative 30 μM tetrodotoxin (TTX)-sensitive and -resistant Na⁺-current traces in left ventricular (LV) myocytes and summarized data on cell capacitance (Cm) and peak and late currents in control (15 cells from 6 CTL rabbits) and electrical storm (7 cells from 3 ES rabbits). Data comparison between groups including multiple cell measurements per animal was performed with mixed effects model (multilevel). *P<0.05, ***P<0.005 vs. CTL. (C) Immunoblots of auto-phosphorylated and oxidized Ca²⁺/calmodulin-dependent protein kinase II (p-CaMKII and ox-CaMKII) and CaMKIIδ and overall data on band intensities in CTL (n=6), non-ES (n=5) and ES (n=10) LV tissues. *P<0.05 vs. CTL. (D) Immunoblots of phospho-Na_V1.5 at S571 (Na_V1.5-P571), total Nav1.5 and Nav1.8 and overall data on band intensities in CTL (n=6), non-ES (n=5) and ES (n=8) LV tissues. Group data were compared using Kruskal-Wallis ANOVA followed by Dunn tests. *P<0.05, ***P<0.005 vs. CTL.

Molecular mechanisms of I_Na-L enhancement in ES rabbits

Recent studies have demonstrated several potential mechanisms contributing to I_Na-L enhancement: Na_V1.5 hyperphosphorylation by CaMKII,⁹ phosphoinositide 3-kinase (PI3K) signaling pathway inhibition,⁸ and neuronal Na⁺-channel Na_V1.8 upregulation.^{10, 11} To assess the roles of these mechanisms, we analyzed key signaling components in the LV tissues.

Figure 6C shows immunoblots for auto-phosphorylated and oxidized CaMKII (p- and ox-CaMKII), and CaMKIIδ. Both p- and ox-CaMKII expression were significantly increased by ≈200% in ES and by ≈90% in non-ES, respectively, without significant change in total CaMKIIδ expression. Fractional phosphorylation and oxidation of CaMKII were both ≈2 times greater in ES than in non-ES. Immunoblots for Na_V1.5, Na_V1.5-P571 (phosphorylation site by CaMKII) and Na_V1.8 are shown in Figure 6D. Na_V1.5-P571 expression was increased by ≈130%, but the fractional phosphorylation (Na_V1.5-P571/total Na_V1.5) change was not significantly different in ES. Na_V1.8 expression was greatly increased (by ≈200%) in ES. Immunoblots for total/phosphorylated Akt, a downstream effector of PI3K signaling showed an increase in fractional phosphorylation of Akt in ES by 39% (Figure SIV), indicating pathway activation rather than inhibition.

Nav1.8 expression in myocardium

To investigate the distribution of upregulated Na_V1.8, we performed immunohistochemistry in transverse tissue sections at a level corresponding to the refractory island confirmed by optical mapping in ES #1 heart. Na_V1.8 was strongly expressed at the basal, but less at the apical myocardium of the ES rabbit LV-tissues (Figure 7A). Na_V1.8 expression was weak in CTL or Non-ES #2 LV myocardium (Figure 7B). The overall data showed that the median of Na_V1.8-staining score we defined (0 to 4, see Supplemental Material) was 3.5 (IQR: 3 to 4) at the base, a value higher than at the apex (1.5; IQR: 1 to 2) in ES, which in turn was comparable to values at both the base (1.0 [IQR: 1 to 2]) and the apex (1.0 [IQR: 1 to 1.25]) for non-ES and CTL tissues, which were similar to each other (Figure 7C). Na_V1.8 fluorescence-labeling was visible within ES #1 myocardial cells although the cellular localization patterns could not be determined (Figure 7D). Na_V1.8-positive nerves were present similarly in ES, non-ES and CTL tissues (Figure SV). These findings suggest that Na_V1.8 upregulation in myocardium, but not intracardiac nerve proliferation or redistribution, contributes to I_Na-L enhancement.

Figure 7. Nav1.8 immunohistochemistry.

(A and B) Representative images of immunohistochemistry for Na_V1.8 in left ventricular (LV) tissues. Cross-sections of the LV base and apex in an electrical storm (ES), a control (CTL) and a non-ES heart. (C) Overall data on the Na_V1.8-staininig score in ES (n=4) and a combined group (n=6) comprising 3 non-ES (gray) and 3 CTL (white-filled circles) rabbit LV tissues. Group differences were evaluated using Kruskal-Wallis ANOVA followed by Wilcoxon signed-rank tests. *P<0.05, **P<0.01 vs. ES base. (D) Na_V1.8 fluorescence-labeling in ES #1 and CTL tissues. Bar scale indicates 100 μm.

To study a biological correlate, we tested the effects of a selective Na_V1.8 blocker, A-803467 (10 mg/kg i.v.), on epinephrine-evoked TdP. Figure 8A shows representative ECG monitoring in an ES rabbit with 5 VF episodes. In the rabbit with a slow escape rhythm and an isolated PVC at baseline, epinephrine infusion provoked PVCs, short VT-runs and TdP, which were completely suppressed by A-803467. The QT-interval prolongation caused by epinephrine was reversed by the compound. Overall data from 4 ES rabbits showed that A-803467 reduced median QTc-interval in the presence of epinephrine infusion from 230 ms (IQR: 227 to 268 ms) to 178 ms (IQR: 166 to 234 ms) (P<0.05) and suppressed epinephrine-evoked TdP events (P<0.05) without any change in QRS-width (Figure 8B and 8C).

Figure 8. Effects of intravenous A-803467 injection.

(A) ECG monitoring at baseline, during epinephrine infusion followed by A-803467 10 mg/kg injection in an ES rabbit (day 64). Values indicate RR- and QT-intervals. (B and C) Overall data on QTc, QRS-interval and the total numbers of premature ventricular complexes (PVCs), short VT-runs (<6 beats) and Torsades des Pointes (TdP, ≥6 beats) (n=4). Friedman ANOVA followed by Dunn tests was used for group comparison. ^#P<0.05 vs. epinephrine.

Discussion

The electrophysiological substrate for ES associated with QT-prolongation has been explored in animals. ES was associated with island-like long APD regions at the LV with increased spatial APD-dispersion, which were exacerbated by epinephrine. We detected drifting rotational activity in the periphery of the refractory island during TdP and found correlation of APD-dispersion with the number of VF episodes. The localized APD prolongation was attributed to I_Na-L enhancement probably due to upregulation of the neuronal Na⁺-channel Na_V1.8 within myocardium, the notion supported by in vivo antiarrhythmic actions of A-803467.

Relationship to previous studies

The cellular and molecular bases of arrhythmogenic ventricular remodeling in CAVB rabbits have been characterized to some extent previously.^12–14 CAVB rabbit cardiomyocytes show APD prolongation with concurrent early and delayed afterdepolarization (EAD/DAD). Downregulation of subunits encoding rapid and slow delayed-rectifier K⁺-current components (I_Kr and I_Ks) underlies APD prolongation.^{12, 13} CaMKII activation and Ca²⁺-handling alterations promote arrhythmogenic EAD/DAD.¹⁴ CAVB rabbits with ICDs develop ES in association with CaMKII activation and phosphorylation abnormalities of Ca²⁺-handling proteins.¹ From these findings, we assumed a predominant role of EAD/DAD-related triggered activity versus reentry as mechanisms of VT/VF in this model. In the present study, however, we observed characteristic spatial APD-dispersion enhancement, detected drifting rotational activity responsible for TdP and found correlation of APD-dispersion ex vivo with the number of VF episodes in ES rabbits in vivo. These findings suggest that the substrate for TdP-driving rotors importantly contributes to degeneration into VF and supports the mechanism described by Jalife and colleagues,¹⁵ whereby either sustained VT or VF results from rapid rotational activity. It is likely that TdP-driving rotor formation facilitates VF occurrence through conversion into stabilized rotors with high frequency or by disorganization of the activity emanating from the rotors. We conclude that EAD/DAD-mediated triggered activity constructs PVCs which initiate reentrant excitation followed by drifting rotors, resulting in TdP perpetuation, and that the antiarrhythmic action of a calmodulin antagonist W-7 reported previously¹ was likely exerted via suppressing PVCs.

Na_V1.8, encoded by SCN10A, is strongly expressed in nociceptive sensory neurons of the dorsal root ganglia and cranial sensory ganglia and is present in cardiac ventricular tissues at lower abundance than Na_V1.5 (SCN5A)⁷ A series of work following genome-wide association studies showed evidence for Na_V1.8-mediated I_Na-L in mouse and rabbit cardiomyocytes perfused with the selective Na_V1.8 blocker A-803467 and SCN10A^−/− mice.^{7, 11, 16} Our findings in this model, together with evidence of Na_V1.8 upregulation in human hypertrophied and failing hearts,^{10, 11} suggest that dysregulated Na_V1.8 subunit-related channel activity may be one of the arrhythmogenic factors in cardiac disorders.

Novel findings and potential significance

This is the first study to characterize the electrophysiological substrate for ES associated with QT-prolongation in a clinically relevant chronic animal model and to present evidence suggesting that reentrant activity contributes to TdP perpetuation in this not-uncommon form of ES. Some investigators have reported a meandering spiral wave in rabbit hearts subjected to hypokalemia plus quinidine¹⁷ or dofetilide¹⁸ and in a transgenic long-QT syndrome type-2 rabbit model.¹⁹ Our detection of a clockwise, counterclockwise, figure-of-8 and dual rotor, all of whose cores were located in the periphery of the island, suggests 1 or 2 scroll source(s) with various types of filaments, potentially contributing to undulating QRS-axis formation. Furthermore, our findings that rotors were anchored by the island were supported by data obtained from the accumulation map of phase-variance that we developed to identify PS distribution: high phase-variance values were distributed in a shape almost identical to the edge of the island.

Limitations

Our ability to map electrical activity was restricted to the epicardial surface. Data on the 3D-structure and its modulation by epinephrine or I_Na-L inhibitors cannot be provided in our model. The size and/or shape of the 3D-island may be related to the severity of arrhythmia indices. There was the wide dispersion of the band intensities of Na_V1.8 immunoblotting. We believe that it was because pieces of samples from the entire LV free wall tissue were studied rather than restricting sampling to the island or LV base. The dosage of A-803467 (10 mg/kg) used in this study may be high, and therefore it cannot be excluded that the antiarrhythmic action was exerted via blocking Na_V1.5. To provide definitive proof-of-concept that Na_V1.8 activity is central to the mechanism underlying VT/VF in this model, future studies are needed to assess efficacy of in vivo RNA interference-mediated Na_V1.8 knockdown.

Conclusions

A tissue-island with prolonged refractoriness created by increased I_Na-L contributes to the generation of drifting rotors that underlies ES associated with QT-prolongation in this model. Na_V1.8 likely underlies torsadogenic I_Na-L and is an interesting candidate for new-drug targeting against ES.