CAVB rabbits show biventricular hypertrophy due to sustained bradycardia and QT prolongation, most of which develop abnormal QTU complex and spontaneous TdP and die suddenly, the electrophysiological features similar to human LQTS (31, 32). When instrumented with implantable cardioverter-defibrillators (ICDs) and followed up for approximately 100 days, the animals have repetitive TdP as non-sustained VT and VF episodes. Electrical storm characterized by clustered VF episodes occurs in ≈50% (Figure 1) (33).

We have recently demonstrated in a perfused heart from an electrical storm rabbit that drifting rotors anchored by a refractory island drive TdP (7). The findings obtained from this animal model are briefly mentioned here with the illustrated summary (Figure 2). a.

Electrical storm hearts had an island-like regions with long APD at the LV base, leading to increased spatial APD dispersion, which were exacerbated by epinephrine (Figure 2A).

Na_V1.8-mediated I_Na−L has been implicated as the substrate for TdP in this model. High concentration tetrodotoxin (30 μM)-resistant I_Na−L was increased in the LV myocytes and Na_V1.8 was upregulated in the LV tissues and expressed within myocardium corresponding to the island location in optically mapped electrical storm rabbit hearts, whereas Na_V1.8-positive nerves were present similarly at the base and apex in electrical storm rabbit tissues (7). Indeed, the selective Na_V1.8 blocker A-803467 (10 mg/kg, i.v.) attenuated QT prolongation and suppressed TdP evoked by epinephrine in electrical storm rabbits (7). We show here 2 cases supporting the agent to possess antiarrhythmic actions. In a rabbit with 2 VF episodes, epinephrine infusion evoked repetitive TdP pause-dependently when the pacing was turned off, a provocative measure that was ineffective after A-803467 injection (Figure 4A). In another rabbit, epinephrine-reinfusion after treatment with A-803467 did not evoke any ventricular arrhythmia (Figure 4B).

Na_V1.8 encoded by SCN10A is strongly expressed in nociceptive sensory neurons of the dorsal root ganglia and cranial sensory ganglia and much less abundant in hearts than Na_V1.5 by SCN5A that mediates cardiac conduction velocity (36). A series of work following genome-wide association studies provides 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 (36, 37). Na_V1.8 is upregulated in human hypertrophied and failing hearts and A-803467 reduces I_Na−L, abbreviates APDs and reduces Ca²⁺-spark frequencies and DADs (38, 39). These findings suggest that Na_V1.8 blockade is a more attractive strategy for treating ventricular tachyarrhythmias than Na_V1.5-selective I_Na−L blockers, preventing arrhythmia without relevant effects on peak I_Na. Sossalla and colleague has recently reported that Na_V1.8-driven I_Na−L is CaMKII-dependent in human failing cardiomyocytes (40). Since electrical storm is associated with striking CaMKII activation (33), it would be interesting to assess involvement of Na_V1.8 phosphorylation by CaMKII in enhanced I_Na−L in our model.

Some investigators provide mechanistic insights into Na_V1.8 interaction with Na_V1.5 in cardiomyocytes. Na_V1.8 functions as a transcription factor for Na_V1.5. The C-terminal portion of Na_V1.8 encoded by SCN10A-short transcripts which appears to be genetic variants in and around SCN10A, increases Na_V1.5 expression and channel current in mouse cardiomyocytes (41). In culture cells, the Na⁺ current increases 2-fold in HEK cells coexpressing wild type SCN10A and SCN5A, compared to those expressing SCN5A alone, and the Na⁺ current decreases in HEK cells coexpressing SCN10A with Brugada mutation and wild type SCN5A (42). Further studies are needed in our model to investigate whether Na_V1.8 overexpressed in electrical storm hearts is a full-length or a C-terminal portion.

While the potential targeting of Na_V1.8 against ventricular arrhythmias in failing and hypertrophic hearts is intriguing, the expression and function of Na_V1.8 in structurally normal hearts, where TdP can occur, remain debatable. More studies would be highly required to clarify whether and how Na_V1.8 blockade with A-804367 affects peak and late components of Na⁺ currents in myocytes from electrical storm rabbits and other models of LQTS.

Spatial dispersion of repolarization has been implicated as a substrate of TdP in CAVB dogs. In studies by Vos and colleagues (46, 47), using 56 needle electrodes distributed transmurally to record 224 unipolar electrograms simultaneously, TdP was initiated at the site with maximal heterogeneity of repolarization in the LV and the selective I_Na−L inhibitor GS967 abolishes TdP through reduced spatial dispersion of repolarization. These findings in CAVB dogs are supported by our results showing that the first beat of TdP following a triggering PVC was a reentrant wave arising in the periphery of the long APD island (Figure 2). In addition, Vos's group originally reported in CAVB dogs that more than 90% of the beats during TdP were focal (48), but a most recent study by his collaborators has developed a novel methodology to detect reentry loops by using graph search algorithms and elegantly demonstrated that TdP can be driven by focal activity as well as reentry depending on the duration of the episode: longer-lasting TdP is maintained by reentry (Figure 5A) (43). Nonterminal TdP that lasts >10 s was all perpetuated by reentry.

A mixture of focal activity and reentry has been reported to underlie EAD-mediated arrhythmias in rat/rabbit hearts exposed to H₂O₂ (49) or hypokalemia in the absence or presence of dofetilide (50) and a transgenic LQT2 rabbit model (51) as well as in the rabbit ventricle simulation model (52). Repetitive focal activations from single or multiple foci from EAD zones, together with regional APD gradients by areas with EADs next to regions without EADs leading to localized conduction block and initiation of reentry, account for the combination of mechanisms underlying EAD-induced TdP and polymorphic VT (53). Moreover, recent simulation studies (54), together with experiments in transgenic LQT2 rabbit hearts (51) and normal rabbit hearts during conditions with I_Kr blockade with E4031 (0.5 μM) plus 50% K⁺/Mg²⁺ (44), have strengthened the idea that steep APD gradients rather than EADs give rise to both the triggering PVC and the substrate for reentrant tachyarrhythmia, the mechanism termed “R-from-T” (Figure 5B) (54). The TdP event we detected in an electrical storm rabbit heart was triggered by a PVC via classical “R-on-T” mechanism in which an exogenous PVC encounters a refractory region (Figure 2), but not “R-from-T”. Since the number of tachyarrhythmia events optically mapped and analyzable was limited, we could not detect ventricular tachyarrhythmias with the “R-from-T” initiation mode. Some PVCs certainly originated at regions adjacent to the long APD island in electrical storm hearts, suggesting that “R-from-T” may be involved in TdP in this model. However, roles of “R-from-T” vs. “R-on-T” in arrhythmogenicity remain unclear.

There are differences in the observed phenomena at the onset and maintenance of arrhythmias between rabbit models. Studies in normal rabbit hearts subjected to the condition of QT interval prolongation by bradycardia, I_Kr blockade with E-4031 or dofetilide and reduced K⁺/Mg²⁺ showed that bursts of EADs coupled to Ca²⁺ oscillations near the base (area of long APD) occurring in zones as small as 1 mm² correspond to TdP (Figure 5C) (11, 55). In electrical storm rabbit hearts, EADs were observed in 5/14 electrical storm hearts and some PVCs occurred from phase 2 of the preceding action potential in electrical storm hearts, but neither reentrant excitation wavefronts propagating from the phase 2 PVC nor a burst of EADs were detected. Membrane voltage oscillation resulting from multiple foci in a non-electrical storm rabbit heart with homogeneous APD prolongation (Figure 3) is an etiology different from bursts of EADs observed in those models (11, 55). The high concentration (0.5 μM) of E-4031 or dofetilide used in the acute model produces prominent APD prolongation, to an extent much greater than that occurring spontaneously in our chronic model, possibly explaining some of the discrepancy. A recently published study in a guinea-pig surrogate model of LQT3 demonstrating that voltage/calcium uncoupling predates sustained ventricular tachyarrhythmia provides evidence of reentrant excitation responsible for TdP potentially degenerating into VF (56).

Spatially heterogeneous prolongation of activation recovery interval (ARI), a surrogate marker of APD, of ventricular epicardium has been demonstrated in genotyped LQTS patients, using non-invasive ECG imaging (ECGi) technique (25). Repolarization gradients resulting from the localized ARI prolongation are found steeper in patients with than without history of syncope or sudden cardiac arrest and ICD-defibrillation (Figure 5D). However, ionic mechanisms of heterogeneous ARI prolongation in congenital LQTS, the syndrome monogenic but with variable phenotypes, remain unclear. The findings in electrical storm rabbits agree with the human data: the boundary of the long APD island likely corresponds to regions with steep repolarization gradients. Our proposal of Na_V1.8-mediated I_Na−L enhancement provides insights into the pathophysiology of LQTS. Acquired regulation/modification of cardiac Na⁺-channels by Na_V1.8 may be one of the important factors involved in the heterogeneous substrate in LQTS.

Studies in simulation models have reported that the polymorphism of VT in the ECG, including TdP, can be determined by a Doppler effect resulting from drift of a single spiral wave (28–30). In agreement with the modeling studies, some investigators reported a meandering spiral wave in normal rabbit hearts subjected to hypokalemia plus quinidine (27) or dofetilide (45) and in a transgenic LQTS type-2 rabbit model (57), but rotational activation was detected in only a few beats during TdP (Figure 5E) and none one has characterized the dynamics of spiral waves. Our study in electrical storm rabbits is the first to report drifting rotors in a unique manner. The detection of a clockwise, counterclockwise, figure-of-8 and dual rotor, all whose cores swirl in the periphery of the long APD island, suggests 1 or 2 scroll source(s) with various types of filaments, potentially contributing to undulating QRS axis formation.