Advanced Electroanatomic Mapping: Current and Emerging Approaches

At their core, all EAMs provide an accurate and stable representation of the anatomy of the chamber of interest. Gating for respiratory and cardiac motion further stabilizes catheter localization within chambers. Other important structures such as the His bundle, fascicles, and course of the phrenic nerve can be tagged on the anatomic shell. Merging of echocardiographic and other imaging can localize structures such as the esophagus, aortic and pulmonary leaflets, origins of the coronary arteries, and important arrhythmogenic structures such as the papillary muscles and the moderator band. Some systems (Table 1) include modules that can superimpose the coronary arteries to improve safety of epicardial ablation.

A reliable voltage map displaying peak-to-peak electrogram amplitude is cardinal to defining traditional anatomic concepts of substrate—low-voltage regions representing structural abnormalities of fibrosis or scar. Multipolar catheters have greatly facilitated rapid acquisition of large number of points. However, inadequate catheter contact and catheter-induced ectopy remain impediments. Although parameters have been defined for normal and abnormal voltages in the various chambers, it should be recognized that scar tissue can be easily obscured by thin layers of normal endocardium. For example, using a 3.5-mm-tip electrode to a 1-mm ring electrode, a bipolar electrogram amplitude > 1.5 mV is observed in > 95% of sites in the normal ventricular endocardium [12••]. Amplitude < 1.5 mV is associated with fibrosis occupying > 75% of adjacent LV wall thickness. However, normal endocardium overlying fibrosis can generate a > 1.5 mV amplitude signal especially in the non-ischemic cardiomyopathies. Thus, while the < 1.5 mV threshold is relatively specific for fibrosis, its sensitivity is limited. In addition, use of smaller electrodes increases the amplitude of local electrograms relative to remote low-voltage regions [13•]. The ability to alter voltage scale ranges and display of unipolar voltages for a “far-field” view can be helpful in detection of intramural substrates [14].

Arrhythmia mapping relies on defining activation during an ongoing tachycardia or, if a tachycardia is non-inducible or hemodynamically unstable, definition of its substrate using features that may be implicated in its maintenance. Most scar-related ventricular tachycardia (VT) are mapped and ablated using this latter strategy as 80% of induced VTs can be hemodynamically unstable.

Activation mapping can point to an arrhythmia mechanism. Focal or micro-rentrant tachycardias usually have a centrifugal spread of activation from their exit. Macro-rentrant tachycardias demonstrate continuous activation. Detection requires a window of interest typically equal to the tachycardia cycle length and a clearly identifiable point on the QRS or an intracardiac electrogram selected as the fiducial for timing reference. For focal arrhythmias, it is desirable to have the beginning of the window of interest start shortly before the anticipated time of activation from the focus, such that the rest of the chamber is later. With macroreentry circuits, recordings should encompass a region where activation coded earliest meets the activation that is coded the latest and represent the majority (> 85–90%) of the circuit. Inability to define a complete reentry circuit may indicate incorrect annotation of electrograms, due to limited spatial sampling in the chamber or that part of the reentry circuit is not accessible (e.g., intramural, epicardial, or in an adjacent chamber), or may in fact challenge the mechanism.

Substrate mapping for arrhythmias relies on the detection of potentially arrhythmogenic features. It is important to note that few features are pathognomic for an arrhythmia. Features for VT include low-amplitude fractionated electrograms within scar regions indicating viable conducting channels, late potentials due to delayed conduction, and demonstration of tissue decoupling with pacing (decrement-evoked potentials (DEEP) [15]) or late abnormal ventricular activity (LAVA) [8] (see below). Delayed conduction can be defined by pacing from suspected sites and demonstrating long stimulus to QRS delays (> 40 ms). If the paced QRS approximates VT morphology, its affords further support for the site of pacing in close proximity to a critical isthmus. Potential features for supraventricular arrhythmias and AF include fractionated signals, low voltage representing fibrosis or scar, isochronal crowding suggesting slow conduction at a presumed critical region, and others.

Current EAMs import precise localization of ablation lesions. With the aid of automated lesion tagging modules based on catheter stability and other parameters, lesion sizing and inter-lesion distances can be used to assess gaps in ablations that aim to create a line of conduction block. The advent of contact force sensing has allowed creation of ablation indices (AIs) that incorporate time, power, and contact force into a single index. In addition, ablation tags can be customized and color coded to include catheter stability, and impedance drops. Proprietary ablation indices (lesion index (LSI) for Ensite Precision and force time integral (FTI) for Carto) correlate with histological lesion formation [16] and can be used to guide ablation for AF [17]. Their utility in ventricular tissue is less well defined. In an ovine model, AI appeared to correlate better with volume than lesion depth [18]. In this study, AI did not correlate with lesion size when half normal saline rather than normal saline was used for open irrigation. Gasperetti et al. used an AI value of 490 for the RVOT free wall and 610 for the RVOT septum with good results and no complications [19].

While ablation indices are useful as a rough guide to lesion creation, lesion durability is less well defined due to wide variability in wall thickness [20], tissue remodeling, disease progression, and other factors. Several considerations, such as the lowest AI that would prevent collateral damage to the esophagus, are undefined. Alternative energy forms such as pulse field ablation with a high threshold for non-cardiac tissue provide one likely solution [21].

High-density mapping of the triangle of Koch has sought to further define the anatomy and physiology of the slow pathway area to facilitate ablation for AV nodal re-entrant tachycardia. Single-center studies describe a low-voltage bridge (LVB) that can be discerned in a field of high voltage between base (coronary sinus ostium) and apex of the Koch’s triangle. Two patterns may be observed (Fig. 2). Ablation at the site of the bridge has a high chance of suppression of slow pathway function. However, the successful site may still be 2–4 mm away from the point of the LVB in 50% of patients [25]. Some investigators have described the LVB as a non-specific finding present in patients without AV nodal re-entry.

Other novel tools have been incorporated into electroanatomic mapping to assist in ablation of AVNRT and other arrhythmias. Ripple mapping displays peak-to-peak voltages, rather than activation times, as a movie of 3D bars changing in length as a function of their voltage–time relationships emerging from the anatomical surface. In Koch’s triangle, such ripples [26] can reveal prolonged, low-amplitude signals which helped identify sites for cryoablation in a pediatric population with fewer lesions than conventional mapping.