Finding the entrance before you can exit

Scar-related, re-entrant, ventricular tachycardia (VT) is a common clinical problem. It arises in the human heart that develops fibrosis and scar tissue. It is encountered postinfarction but also complicates various cardiomyopathies, forms of structural heart disease and may occur after cardiac surgery. Mechanistically, most scar-related VT is related to a reentrant circuit maintained by a constrained, narrow isthmus within scar that harbors slow diastolic activation. Entrance and exit sites, to and from the isthmus, tend to reside at the border zone between scar and healthy tissue.1 What is less well understood are the histological and electrophysiological characteristics of abnormal myocardial tissues that deem them to become entrances, exits, or isthmus sites that eventually maintain scar-related VT, leading to recurrent episodes and their possible sequelae clinically. There is a long-held debate about whether these properties are fixed and related primarily to abnormal ventricular histology or functional and related to a plethora of often stochastic electrophysiological changes.2 Recently, there have been studies to support both arguments. On the one hand, for example, Nishimura et al. recently reported, in Circulation, that up to 84% of constrained isthmus sites are bordered by unexcitable scar present (if sought for hard enough) even in sinus rhythm. Fixed lines of block delineating an isthmus is an attractive hypothesis as it suggests they can be mapped in multiple rhythms to facilitate catheter ablation of scar-related VT beyond targeting the diastolic isthmus during tachycardia.3 On the other hand, multiple groups have reported the functional nature of these lines of block that eventually constrain a narrow conducting isthmus. For example, Ciaccio and colleagues have reported on the importance of functional lines of block arising owing to source— sink current mismatches at areas of tissue thickness changes. Despite these being lines of functional block, their signatures can also be identified during sinus rhythm.4, 5 In this issue of the Journal of Cardiovascular Electrophysiology Hayashi and colleagues present thought-provoking data to add to the debate between functional and fixed lines of block in the circuit of scar-related VT. The authors studied patients with scar-related VT arising from ischemic cardiomyopathy (ICM) or following correction for tetralogy of Fallot undergoing open heart surgery to perform cryoablation to control recurrent VT episodes. Scar-related VT was mapped using an electrode array mounted onto a latex balloon that was sutured to the endocardium and another sock-like array sutured to the epicardium of both ventricles. Nine VT circuits were mapped in seven patients, five circuits arising in patients with ICM and four circuits in patients with corrected tetralogy. The benefit of the balloon and sock arrays is twofold: first, the entirety of the chamber is mapped simultaneously and second, it is possible to map a single beat of VT. The authors analysed conducted VTs when at least five beats had been successfully mapped using the arrays. What is thought-provoking is the authors' decision to study spontaneous termination of scar-related VT using their intraoperative mapping approach. As the authors point out, spontaneous termination has not been studied in detail previously. There are four main studies which describe the spontaneous termination of scar-related VT that we are aware of, and only one was performed in humans. All were performed in the 1990s.6-9 Conceptually, the idea of mapping spontaneous episodes of conduction block and, therefore, termination is an intriguing and novel approach to gaining also occur during some portion of the QRS insight into the substrate of scar-related VT. It is certainly worth revisiting the nature of spontaneous scar-related VT termination using modern mapping approaches in human hearts. In their paper, Hayashi and colleagues observe that spontaneous termination of VT mainly occurs at the entrance to the isthmus. They demonstrate that 89% of VT's mapped spontaneously terminated during late systole, or early diastole, at the VT circuits entrance into the isthmus. Only 11% of circuits terminated in the isthmus itself, and none at the VT's exit site. To determine the timing of termination, the authors arbitrarily split the reentrant circuit cycle into two windows: systolic (QRS) and diastolic (between QRS). The diastolic window is further subdivided into thirds as diastolic windows representing the entrance into the isthmus, the mid-isthmus and the exit from the isthmus. Of course, entrance and exit into and from the circuit will also occur during some portion of the QRS so the window's systolic component was also divided into thirds. Their method also uses annotation of bipolar (instead of unipolar) electrograms to its first local deflection to determine a local activation time (LAT) to create a 3D activation map of the VT. The anatomical location of termination and its approximation to an entrance site was also determined by analysis of these maps. Furthermore, the authors demonstrate that cycle length changes do not predict spontaneous VT termination, which could be preceded by slowing, acceleration, or even abrupt stopping. Nonetheless, the activation pattern of VT did change before termination in all circuits mapped. The data presented by Hayashi and colleagues challenge the few previous studies investigating spontaneous termination of scar-related VT. In the seminal study by Marchlinski and Callans, in humans with chronic infarct-related VT, cycle length variability (greater in patients on antiarrhythmic medications) was noted before spontaneous termination.7 No pattern of variability, however, predicted termination. In canine studies, spontaneous termination was noted to occur in the isthmus, not the entrance, whilst cycle length variability was frequently but not always noted (i.e., abrupt termination was also seen).6, 8, 9 What is important to note is that sites of entrance termination showed significant local electrogram heterogeneity in the lead-up to termination, with 44% showing beat-to-beat alternans in the entrance site but only 1.8% showing such change in the isthmus. Furthermore, entrance sites also showed significant decrementation after extra-stimulation during sinus rhythm. These properties are important as they suggest local conduction heterogeneities are also present in sinus rhythm, possibly owing to source-sink mismatches in current density. Recently, multiple studies have investigated the substrate of scarrelated VT as mapped during sinus rhythm. Two approaches, decrement evoked potential (DEEP) mapping and isochronal late activation mapping (ILAM), seem to elucidate areas of slow and decelerating conduction during sinus rhythm.10, 11 These areas are thought to correlate well with the entrance and isthmus of scar-related VT and seem to provide successful ablation targets. “Channel identification” based approaches to mapping the substrate of scar-related VT remain important and are associated with successful ablation outcomes.12 however they appear to provide less functional information about the circuit and substrate. Interconnected channels are (theoretically) helpful in identifying the VT's exit and isthmus, and less so the entrance, whose importance is highlighted by Hayashi et al. The paper is not without limitations, however. As the authors acknowledge, their sample size is small. Their five-beat inclusion criterion inevitably leads to the inclusion of nonsustained VT, but we are reassured that this resembled a clinically encountered morphology when included. In our opinion, it would have been beneficial to include a voltage amplitude analysis, or map, of the bipolar electrograms acquired. This would be particularly interesting as it would help better define the areas of possible source-sink mismatch that led to a functional block. Furthermore, annotating LAT to bipolar electrograms can be problematic in identifying the true instance of local tissue activation. Despite the paper's strengths, it is impossible to use it as definitive evidence of the mechanism of spontaneous conduction block and the termination of scar-related VT. Even so, the data presented by Hayashi and colleagues in this issue of the Journal question certain fundamentals about the mechanism of scar-related VT. They implicate the circuit's entrance site as a critical reentry area and question the fascination with the VT exit site mechanistically and as a target for ablation, be it during VT, through pace-mapping, or using substrate equivalents. The authors are right to wonder how their data question current ablation targets for scar-related VT, and we wonder whether their data provide a starting point for revisiting contemporary strategies for VT ablation. The authors declare no conflicts of interest. Disclosures: None.