Atrial fibrillation substrate mapping with emphasis on voltage-based guidance

Abstract

Voltage substrate mapping is a promising tool for the treatment of atrial fibrillation (AF). It is helpful to detect atrial fibrosis, which includes areas with low bipolar voltage, heterogeneous conduction properties, and shortened effective refractory period. The voltage amplitude is typically defined as the maximal peak-to-peak level within a specified time window of interest. Contemporary electroanatomic mapping platforms now enable many thousands of data points to be mapped, so that a geometric model of the atrial endocardium is constructable over a short period of time. This mapping procedure is often done with bipolar electrodes to cancel the far-field signal. The recording site coordinates are projected onto an atrial shell, with interpolation of the voltage data across the shell surface. The amplitude of the recorded bipolar electrogram depicted on the three-dimensional shell provides detailed information for substrate mapping. Wherever there are areas of low peak-to-peak voltage, it is thought to mark the presence of abnormal tissue properties and conduction. However, uncontrolled variables and environmental factors affecting voltage level include the oncoming electrical activation wavefront direction, the catheter incidence angle, the force applied to the catheter, and the region-variable shape and structure of atrial tissue. Techniques and settings to acquire atrial voltage data for AF analysis have not been standardized. Methods to characterize atrial electrograms are also presently limited. These factors affect quality and reproducibility of the mapping results. Herein, voltage substrate mapping and its variables pertaining to AF and radiofrequency ablation are described and discussed, with suggestions for future work efforts.

Key Words: Ablation; Atrial fibrillation; Sinus rhythm; Substrate mapping; Voltage

Core Tip: Voltage mapping is commonly used during substrate mapping to detect target areas for ablation in patients with atrial fibrillation (AF). Analysis of the substrate may take the form of detecting low voltage areas or the average voltage over an entire region. Often, areas which are targets for ablation are located where the voltage level is lower than normal, indicating disease and damage to the substrate. Furthermore, as AF progresses from paroxysmal to persistent type, the voltage level of the left atrium tends to diminish. This is often associated with atrial remodeling, in which structural and electrical changes occur during arrhythmia. Protocols for finding target areas using voltage mapping are described, and future research efforts to improve mapping and outcome are discussed.

INTRODUCTION

In this review, aspects of substrate voltage mapping will be described and discussed for the detection and localizing of ablation targets in patients with atrial fibrillation (AF).

LOW VOLTAGE RANGE AND SCAR

Electroanatomic mapping provides information pertaining to local voltage abnormalities that may be used as surrogate markers for the presence of atrial fibrosis and other structural impediments to conduction of the electrical activation wavefront[1]. The implementation of voltage mapping and analysis during electrophysiologic (EP) study enables the in vivo assessment of scar, which can then be targeted with radiofrequency, pulsed field[2], or other ablation energy[3,4]. Low voltage is often a focus of ablation treatment in AF, but its relationship to onset, perpetuation, and termination of the arrhythmia requires further elucidation[5]. A cutoff value for left atrial low bipolar voltage mapping has been described as less than or equal to 0.5 mV, as measured during sinus rhythm[6]. Further, a specific voltage range of 0.2–0.45 mV during sinus rhythm has been reported to accurately delineate left atrial scar distribution[7]. The single optimal cutoff value for detection of low voltage when recording during AF rather than sinus rhythm is 0.2 mV[8]. Although a low voltage threshold or voltage range can thus be utilized to indicate presence of atrial fibrosis, the threshold voltage level defining fibrosis is not necessarily constant, and may diminish with increased atrial structural remodeling[9]. Voltage depression when recording during AF as opposed to sinus rhythm may be caused by functional refractoriness and rapid electrical activation[10]; rhythm type during data acquisition is therefore an important consideration. Although bipolar recordings are often utilized for analysis, unipolar recordings may better characterize left atrial fibrosis by capturing more comprehensive transmural features[11].

Regional voltage differences are also apparent in patients with atrial arrhythmias, which may reflect the presence of altered tissue thickness and variable electrical activation wavefront propagation. The most frequent location of sinus rhythm low-voltage areas (LVA) is at the anterior wall, the septum, and the posterior wall of the left atrium (LA)[12-14]. Posterior LVA is associated with a significant increase in persistent AF recurrence[15]. Substrate voltage levels also depend upon patient characteristics of age, sex, co-occurrence of dilated or hypertrophic cardiomyopathy, the type of AF, left atrial surface area and volume, sinoatrial node dysfunction, and glomerular filtration rate[16]. These parameters correlate with whether LVA are present, and they may be useful in selecting an appropriate strategy for AF ablation[17,18]. Furthermore, it should be noted that left atrial LVAs are associated with a history of stroke and subclinical silent cerebral ischæmia in patients with non-valvular AF[19]. Yet, most AF patients who suffer stroke have no evidence of arrhythmia in temporal proximity, implying that the stroke is mechanically based and operates via a generalized atrial cardiomyopathy, rather than originating from electrical substrate[20-22].

In terms of correlation, statistical significance, and prediction, left atrial voltage is related to the total sinus rhythm activation time throughout the chamber[23]. Moreover, the total sinus rhythm activation time is greater in patients with persistent AF as compared to those with paroxysmal AF[24]. Hence, conduction velocity is likely to be slower in persistent AF, and a slower conduction velocity is consonant with a diminished left atrial voltage. Patients with persistent AF have significantly lower left atrial voltage levels during AF as compared with the paroxysmal type, even after adjustment for differences in indexed left atrial volume[25]. Moreover, LVAs are larger in patients with persistent as compared with paroxysmal AF[16]. Although persistent AF can be terminated by catheter ablation without linear lesioning, some of these cases require straight-line lesions to prevent macro re-entrant atrial tachycardia (AT)[26,27]. Left atrial voltage is an independent predictor of arrhythmia recurrence, whether the LA is mapped during sinus rhythm or during AF[28]. For paroxysmal AF, the existence of left atrial voltage areas less than 0.4 mV in greater than 10% of the total LA surface area is predictive of arrhythmia recurrence following pulmonary vein ablation[29]. In both paroxysmal and persistent AF, the presence of left atrial LVA after pulmonary vein isolation (PVI) predicts AF recurrence[30,31]. Notably however, recurrence of AF after PVI, which is common, may be due to pulmonary vein reconnections, rather than the presence of the LVA themselves[32,33]. Furthermore, LVA ablation has no beneficial impact on 1-year rhythm outcomes[34]. In addition, preprocedural cardiac magnetic resonance imaging-particularly for substrate mapping-can be useful; however, identifying pulmonary vein reconnections remains challenging, even with cardiovascular magnetic resonance (CMR)-gadolinium-enhanced (GE) imaging[35].

AF TRIGGERS AND DRIVERS

AF triggers can be defined as electrical impulses leading to AF onset from a non-fibrillatory rhythm, and AF drivers can be characterized as electrical phenomena critical to the arrhythmia’s perpetuation[36-39]. Several harbingers of AF triggers and drivers exist. LVAs are linked to AF trigger sites[40]. Pulmonary vein triggers are associated with diminished voltage levels in the pulmonary vein antra, while superior vena cava triggers correlate with lower voltages in the right atrium[41]. Non-pulmonary vein triggers arise from left atrial degeneration, which thereby contribute to persistence in AF[42]. Even when there are no LVA, it has been suggested that left atrial electrophysiological function may be alterable to form AF substrate[43]. The delineation of electrogram morphology has also been proposed to identify AF triggers and drivers. Complex fractionated atrial electrograms (CFAEs) are defined as electrograms with a very short cycle length (≤ 120 megaseconds), or fractionated signals exhibiting continuous electric activity[44]. An example of a CFAE from a paroxysmal AF patient is shown in Figure 1, with the temporal variability in electrogram morphology being evident. Regions with CFAE are possible critical target sites for AF ablation procedures[45]. Such CFAEs can occur in regions with slowed conduction or wavefront collision, and they tend to increase with patient age[46]. Furthermore, the dominant frequency (DF), defined as the spectral frequency with the highest power in the physiologic range of interest, often defined as 3–12 Hz[47], has long been considered a driver of the frequency activating the AF substrate. In Figure 2, some AF spectral parameters used in some studies are identified and defined. Most CFAE and high DF sites detected during AF, however, do not correspond with LVA, fractionation, or high DF sites identified during sinus rhythm[48-50]. Hence, these regions likely do not reflect damaged atrial myocardium, in accord with the normal sinus rhythm voltage levels that are observed[50]. Moreover, areas of CFAE and low voltage recorded during AF frequently demonstrate normal atrial conduction properties and electrogram characteristics when the same areas are recorded during sinus rhythm[51]. Therefore, CFAE sites probably do not identify regions with, and are not specific for, abnormal atrial substrate, and both CFAEs and high DF areas during AF presumably do not reflect damage to the atrial myocardium, as is evidenced by their normal sinus rhythm voltage[52]. Instead, the distribution of CFAEs is predominantly along the posterior and high septal RA, where there are marked changes in myocardial fiber orientation. This suggests that the underlying myocardial architecture is a main generator of these waveforms. AF electrogram spatiotemporal dispersion characteristics can be used to guide the catheter to target areas[53,54].

Figure 1 Example of a complex fractionated atrial electrogram from a paroxysmal atrial fibrillation patient. Complex fractionated atrial electrograms are fractionated electrograms comprised of multiple deflections each atrial cycle, caused by the presence of multiple asynchronous electrical activation points. The electrogram deflections often appear to be random, varying in size and shape. Peaks are usually of low amplitude, although larger peaks may be interspersed. The fractionated electrogram recordings tend to differ in patients with paroxysmal vs persistent complex fractionated atrial electrogram, which may reflect the degree of substrate remodeling. Satisfactory quantitative methods to characterize these electrograms are yet lacking.

Figure 2 An example frequency spectrum (top panel) generated from an atrial electrogram which was acquired during atrial fibrillation in a persistent atrial fibrillation patient. The largest spectral peak in the physiologic range of interest, typically 3–12 Hz, which is not a harmonic, is defined as the dominant frequency or dominant frequency. The amplitude or magnitude of that frequency component is labeled the dominant amplitude or dominant amplitude. Two other parameters can be described by first normalizing the range of the ordinate axis to 0–1 mV (bottom panel). The mean value of this normalized frequency spectrum is termed the mean spectral profile (MP), and its SD is labeled the spectral profile (SP). The MP will be larger when the atrial electrogram is more random, such that there is no predominant spectral peak. The SP will tend to increase as the number of spectral components of varying amplitudes increases.

Atrial fibrosis is a demonstrated substrate for AF[3]. The appearance of fibrosis enhances conduction heterogeneity throughout the LA, causing disturbances in the electrical activation wavefront, including delay, conduction block, wave break, and reentry in the atrial wall, all of which may assist in perpetuating arrhythmia[55,56]. Rotational electrical activity, commonly termed rotors, can be meandering or stationary, and they anchor at regions of slow conduction, at gradients in refractoriness, structural discontinuities, and low voltage fibrotic regions[39,57]. A strong correlation has been established between left atrial fibrosis detected via CMR–GE and low endocardial voltage measured during electroanatomic mapping[58]. In AF patients, late gadolinium enhancement (LGE) regions exhibit a pronounced and consistent reduction in unipolar and bipolar voltage compared with non-LGE regions[59]. Left atrial bipolar voltage level is associated with the image attenuation ratio derived from contrast-enhanced multidetector computed tomography[60]. Low atrial endocardial bipolar voltage, as measured during AF, is a commonly used surrogate marker for the presence of atrial fibrosis[61]. The extent of left atrial fibrosis predicts the success of AF ablation at up to five years of follow-up[62]. AF voltage values sampled over several seconds show a strong temporal reproducibility[63], suggesting the duplicability of this measurement. However, of those left atrial targets which result in successful termination, more than half are entirely remote from low voltage areas[64], and in patients with advanced fibrosis, AF ablation has a high failure rate[62]. Thus, AF voltage as a sole parameter has limited value in efficaciously guiding the catheter toward target areas.

ABLATION OF ARRHYTHMOGENIC REGIONS

Catheter ablation, particularly PVI, is an established first-line treatment for paroxysmal and persistent AF[65]. LVA in the LA serves to flag the presence of arrhythmogenic substrate and it is a strong predictor of AF recurrence after PVI in both paroxysmal and persistent AF[66,67]. The extent of LVA is an independent predictor of arrhythmia recurrence, even after substrate homogenization[68,69]. Thus, targeting for ablation sites with distinct activation within or bordering LVAs may yield better outcomes than PVI alone[69]. Multiple studies and meta-analyses have shown that substrate modification targeting LVAs, when added to PVI, results in improved arrhythmia-free survival in patients with persistent or long-standing persistent AF, without increasing procedural risk[70-73]. In one analysis, freedom from atrial arrhythmia at one year was achieved in nearly three-quarters of patients receiving LVA-guided ablation[71]. The SUPPRESS-AF Trial found that LVA-guided ablation decreased recurrence rates compared to PVI alone in persistent AF, supporting such individualized substrate-guided approaches[74]. Ablation of sites with distinct activation characteristics within or at LVA border zones, in addition to PVI, is more effective than a conventional PVI-only strategy for persistent AF[69]. A meta-analysis has shown that ablation of LVAs in addition to PVI reduces arrhythmia recurrence without significant increase in procedure duration, fluoroscopy time, or complication rate[75]. Individualized ablation with high-density left atrial mapping followed by LVA-guided ablation, when LVAs are present, significantly improves outcomes, although at the cost of longer EP procedure time[76]. Such voltage-guided ablation increases one-year AF/AT-free survival in patients when compared to stepwise ablation[3]. In fact, voltage-guided ablation is the only independent predictor of AF/AT-free survival. When AF patients were assigned to LVA isolation vs standard treatment, there was a significantly increased arrhythmia-free survival in the LVA-guided group at one year[77,78]. Patients with LVA have an equally favorable long-term ablation outcome as compared to those without[79]. In both paroxysmal and persistent AF patients who lack LVAs, solo PVI leads to successful outcome[66].

Despite much progress in substrate mapping, AF reappearance rates remain high, ranging to as much as 40% at one-year post-ablation[34]. In patients with paroxysmal AF, LVA ablation beyond PVI may not improve outcome, and the presence of LVAs strongly predicts AF recurrence[80]. Less successful outcomes occurring in persistent AF may result from a longer AF time duration since onset of arrhythmia, which is associated with more extensive substrate remodeling, as well as to the presence of atrial dilatation[81]. Some randomized trials do suggest that the benefit of LVA ablation beyond PVI is inconsistent. Adjunctive ablation strategies following PVI has not consistently improved outcomes in randomized controlled trials[82]. The size and distribution of LVA affect AF initiation, perpetuation, and termination[5]. Furthermore, LVA ablation may not always reduce AF recurrence nor increase AT incidence vs PVI alone[83], despite a connection between LVA determined by voltage mapping in AF and scar observed on CMR–GE imagery[84]. In patients with minimal LVA, it may therefore be sufficient to apply only PVI[66]. When there is extensive atrial remodeling, such as is indicated by the presence of fatty tissue, a predominance of low voltage zones, and/or electrogram fractionation, only then may a more aggressive substrate modification be required[85].

Difficulties in the use of LVA for substrate mapping may result in part from the variability of its characteristics, which can complicate the targeting of arrhythmogenic tissue for ablation[86]. Examples are shown in Figure 3 to distinguish and note the partial but by no means complete repetitiveness of atrial electrogram deflection and amplitude during AF, hence the possibility of variable efficacy in the use of the electrogram for substrate mapping. There is to some extent a linear voltage correlation between sinus rhythm and AF, suggesting a similar extent of left atrial fibrotic substrate may be identifiable during voltage mapping in any rhythm by adjusting the voltage cutoff[87]. When voltage mapping is done in AF, the LVAs are more evident and prominent than in other rhythms[69,87], with a tendency to reveal a larger low voltage extent[6], and indeed AF voltage is superior to sinus rhythm voltage for localizing CMR-GE regions[63].

Figure 3 To show variability of atrial fibrillation electrograms (patient with persistent atrial fibrillation) the frequency spectrum was first computed for electrograms acquired at dual sites in proximity in the left atrium. Similar values of dominant frequency (DF) or DF are evident (left panels), and it is apparent that the DF in each case is the pre-DF–only a few lesser frequency components are evident in each spectrum. Addressing this scenario, the time series was windowed based on the period of each DF, with the resulting segments evident in the right panels. The electrogram deflections for each DF cycle are often overlapping. In the top right panel, the electrogram is broad and biphasic, starting with a negative deflection and ending with a positive deflection. Although there is some cycle-to-cycle variability, the gross morphologic properties appear mostly consistent. The same basic shape and pattern are evident in each cycle, except for a few deviations. The deviations include earlier onset of the early deflection, which corresponds to earlier arrival of the wavefront at the negative pole of the bipolar electrode. The abscissa is given as sample number, and for the digitization rate of 1200 Hz it is an approximation of the time in milliseconds. Likewise, for the overlap of the lower right panel electrogram, there is similarity between segments, although the entire biphasic deflection sometimes occurs earlier in the cycle, suggesting earlier breakthrough from the source or earlier completion of the reentrant circuit, with variable degree of earliness. DF: Dominant frequency.

DISEASE PROGRESSION AND REMODELING

Structural remodeling includes atrial dilatation, fibrosis, and reduced myocardial voltage, and for AF patients, it begins near the pulmonary vein antra and progressively involves other left atrial regions[87-89]. The electroanatomic changes are reflected in reduced regional voltage, slowed conduction velocity, and an increased presence of fractionated electrograms and LVAs. These alterations are more pronounced in patients with persistent vs paroxysmal AF, suggesting that there is a progressive remodeling of substrate as the time since AF onset increases[90]. From the first to second substrate modification procedure, when needed, bipolar mean peak-to-peak voltage in the LA can decrease significantly, with marked increase in LVA surface area[91]. No similar significant changes are evident in the right atrial bipolar mean peak-to-peak voltage or in the LVA surface area during the second procedure, suggesting a predominant role of the LA, but not the right atrium, during disease progression. Electrogram morphology, in terms of the activation wavefront power (as estimated by spectral amplitude) and refractoriness (as estimated by DF) can be utilized to express the severity of disease progression. A schematic is illustrated in Figure 4. The electrogram (top graph) is segmented by windowing. At top, for a window length w = 100 sample points, the first four segments are shown, and similarly for window lengths w = 200 sample points and w = 500 sample points at middle and bottom, which are segmentations of differing window length applied to the original time series signal. For each window length, all segments in the series are then overlapped, and the ensemble mean power is computed and graphed for all segment lengths plotted on the ordinate axis, with the abscissa scale converted to frequency = 1/w. This method of frequency analysis, termed the New Spectral Estimator, can be helpful to use as a comparator to Fourier analysis, and it shows similar levels of significance in the difference in spectral parameters when comparing paroxysmal vs persistent AF[92]. Local bipolar voltage and its remodeled characteristics can in part be surmised from the spectral power, although it is also dependent on electrode size and interelectrode distance, tissue thickness/health, and activation wavefront orientation.

Figure 4 An example atrial fibrillation electrogram is shown in the top graph with variable deflection rate and morphology. The ensemble mean or average of electrogram signal segments of length w can be determined from such a signal. The signal can be segmented using a window width w, where frequency = 1/w. For each w, the signal segments can be summed to compute spectral power at that frequency. This method, an alternate to Fourier analysis, is termed the Novel or New Spectral Estimator. The dominant frequency or dominant frequency so computed (peak frequency) describes the power at the activation rate of the signal. Greater power reflects those larger portions of the substrate in the field of view of the bipolar electrode activating at time intervals w, i.e. the coupling interval or cycle length. Since activation is randomized, it is likely indicative of the local refractory period, with longer coupling intervals portending longer local refractory periods.

During remodeling, a longer AF time duration since onset is also associated with an increased left atrial scar burden, which can negatively impact the success of ablation procedures[93]. However, all patients with AF have lower regional voltages, increased proportions of LVAs, slower conduction velocities, and more complex electrograms as compared to control patients without AF[90]. Since these abnormalities are observed even in the absence of overt structural heart disease, they support the notion of primary electrical remodeling (i.e., electrical remodeling that occurs primarily in response to a functional effect, such as an altered electrical activation sequence)[94,95].

Although prompt termination of paroxysmal AF episodes may temporarily suppress arrhythmia symptoms, it does not halt disease progression, nor does it reverse the underlying progress of atrial remodeling[96]. LVA grade, which is reflective of remodeling, is an independent predictor of paroxysmal AF recurrence following PVI[97]. In a randomized comparison, AF recurred in half of patients who underwent PVI alone, compared to about a third of those who received PVI with additional substrate modification; hence targeting the underlying atrial substrate is appropriate for selected patients[82]. In AF ablation, scar formation as evident on CMR–GE is associated with AF recurrence[98]. However, CMR–GE is currently unsatisfactory in terms of its spatial resolution. Another hurdle as yet to be accomplished is that by integrating imaging with mapping data, a goal as yet unattained, patient stratification might more readily be implemented. Patients with persistent AF demonstrate more advanced remodeling compared to those with paroxysmal AF, including a lower median left atrial voltage, a greater percentage of bipolar voltage regions below a 0.2 mV threshold level, and a higher indexed LA volume (volume/body surface area)[25]. Of note, the shape of the left atrial chamber is predictive of AF ablation failure, independently from and more accurately as compared with atrial volume and AF persistence[99]. Hence, structural properties are of importance in determining arrhythmia persistence. Clinical trials suggest that earlier use of rhythm control, along with AF ablation, is associated with better outcome[100,101], which may indicate the possibility of dampening the progression of remodeling.

Systemic hypertension, especially when accompanied by left ventricular hypertrophy, is another property associated with atrial remodeling[102]. This hypertension is manifested as a global slowing of electrical conduction, regional conduction delay (notably at the crista terminalis), and increased susceptibility to AF induction[102]. While lower global left atrial conduction velocity and widespread voltage reduction are predictive of AF recurrence, the presence of discrete LVAs alone is not independently predictive[103], perhaps indicating the necessity of global property transformation in the remodeling process, to affect change in arrhythmia propensity. The relationship between voltage amplitude and wavefront activation slowing is complex, with some regions of low voltage exhibiting normal activation while other areas with normal voltage demonstrate significant activation slowing[86]. This supports the hypothesis that AF maintenance may also involve diffuse and more globalized substrate remodeling, rather than being dependent on localized LVAs alone. Furthermore, it is known that left atrial voltage reduction is a diffuse process related to the underlying fibrosis pattern, reinforcing the idea that atrial remodeling is not strictly limited to localized, well-defined low-voltage zones[104]. A persistent AF diagnosis, however, does not necessarily indicate the presence of voltage-defined fibrosis, nor does paroxysmal AF indicate a lack of fibrosis[105].

One of the known and hallmark features of electrical remodeling in AF is a reduction in the atrial effective refractory period (ERP), and an impaired ERP adaptation to increased rates, which contributes to local reentry and AF maintenance[106,107]. At the cellular level, AF leads to a marked shortening of action potential duration and impaired rate adaptation. These changes are largely attributed to reductions in the cellular-level calcium current and the transient outward current, thereby altering calcium and potassium handling in atrial myocytes[108,109], which confers remodeling at the microscopic-level.

Epicardial adipose tissue (EAT) remodeling also plays a crucial role in AF development and progression[110]. Patients with AF exhibit a higher volume of EAT and a lower image attenuation value[111]. The volume and attenuation of left atrial-EAT are associated with the presence of LVA within the myocardium. Myocardial bipolar voltage on electroanatomic mapping of the atrium is negatively associated with age and the presence of overlying EAT[112]. Electrogram fractionation in sinus rhythm is positively associated with age, male gender, diabetes, hypertension, and EAT. Electrogram widening is correlated to age, male gender, and EAT. Body mass index is directly associated with the presence of EAT in patients with AF. However, EAT is not a statistical mediator of the association between clinical variables and left atrial scar. Ultimately, lipomatous dysplasia, characterized by excessive fat accumulation in the heart which alters electrical current load and creates reentrant circuit corridors, may contribute to onset of both AF and ventricular tachycardia[113,114].

SOURCES OF VOLTAGE VARIABILITY

Voltage signals recorded for AF analysis are influenced by many factors-including wavefront direction, electrode configuration, tissue properties, activation rate, contact force, and filter settings[115]. There is only a partial correspondence between the voltages obtained in sinus rhythm vs AF-acquired signals, with weaker AF signal recordings being found in patients with persistent AF[87]. Although voltages acquired in both atria are generally lower during AF vs sinus rhythm, there is a greater voltage reduction observed at sites exhibiting the shortest AF cycle lengths (shortest frequencies)[1]. This may be due to the diminished coupling interval, disorganized depolarization, or partial depolarization of atrial tissue. Faster activation rates in general result in larger observed LVAs, particularly in paroxysmal AF[116]. Pacing from the coronary sinus at a rate 50 megaseconds faster than sinus rhythm reduces bipolar voltage and also increases the extent of LVA[117]. The voltage attenuation that occurs with faster pacing rate, which is termed Dynamic Voltage Mapping, demonstrates that left atrial voltage decreases at more than two-thirds of recording sites as coupling intervals shorten[116].

Voltage amplitude is highly influenced by the structural properties of the substrate. It is directly affected by atrial wall thickness, which varies regionally, even in normal atria[14]. Bipolar voltage is significantly higher on the epicardium (mean: 3.05 ± 1.31 mV) as compared to the endocardium (mean: 1.65 ± 0.75 mV)[116]. Increased atrial size and pressure are both associated with a lower atrial voltage[118]. Atrial wall stress and acute atrial dilatation have been linked to the formation of LVAs, particularly in the posterior wall[119]. Impaired mechanical function can result in lower voltage levels[6]. Aberrant atrial structure may cause complex and patient-variable three-dimensional bioelectrical transformations resulting in activation wavefront asynchrony, with observed endo-epi dissociation[120,121]. Hence, endocardial substrate mapping alone may not fully characterize the mechanisms of AF, and sole reliance on endocardial ablation may be insufficient in the treatment of AF.

The data acquisition system utilized to record the electrograms is a major source of variability between investigator teams. Bipolar electrode spacing determines the temporal offset in wavefront arrival, which influences both the signal morphology and the signal amplitude. Greater interelectrode spacing increases electrogram voltage until a plateau is asymptotically reached-typically beyond a 4-mm electrode spacing in healthy tissue-but this plateau is absent in diseased myocardium[122]. Increased bipolar electrode spacing also leads to greater far-field signal contribution. Larger electrodes record higher amplitude signals due to the larger tissue contact area, but these larger electrodes more likely include both normal and diseased myocardium. Adequate contact force is also necessary to ensure reliable voltage recordings. There is a weak correlation between contact force and voltage at low force levels, but no further voltage increase occurs beyond a certain threshold[52]. The filter settings during data acquisition additionally affect electrogram amplitude[123]. Increasing the high-pass filter frequency substantially reduces the amplitude of electrogram deflections. During clinical mapping, typical bipolar filters include high-pass corners between 1–30 Hz and low-pass corners between 300–500 Hz, with the high pass filter in this range significantly altering both electrogram shape and its amplitude.

ELECTRODE POLARITY

Much attention has recently been given to the polarity of the signal recording. Unipolar and bipolar voltage maps are congruous in both sinus rhythm and AF[124], though it is bipolar electrode technology that is ubiquitously utilized for substrate voltage mapping. It is based on the principle that the differential of two separate electrode pole signals results in cancellation of the far-field component. The bipolar signal is formed when the local extracellular signal generated by the activation wavefront arrives at the positive pole of a bipolar electrode, which is added to form a positive deflection in that pole’s signal, and then arrives at the negative pole, which is subtracted by the differential transformer, causing a negative deflection, or vice versa. Thus, a biphasic electrogram shape is obtained when an activation wavefront sequentially propagates past the two poles of a bipolar electrode. If, however, the wavefront leading edge arrives at both poles simultaneously, the near-field as well as the far-field signal may be mostly cancelled, resulting in a low-voltage electrogram output. Hence, voltage level differs for parallel vs perpendicular orientated bipoles with respect to wavefront propagation direction, particularly in sinus rhythm when the orientation of the oncoming wavefront has a tendency to be more constant. Yet due to the randomness and time-variability of recordings made during AF, wavefront orientation then has a less clear effect on bipolar signal voltage[125]. Unipolar electrograms have no such oncoming wavefront orientation bias, but they do not remove atrial far-field activity[39].

Recently, omnipolar technology, which utilizes multiple electrodes to form a differentiated signal, has been introduced[126]. This arrangement can eliminate the wavefront orientation bias evident in bipolar electrode recordings, particularly during sinus rhythm. Omnipolar electrogram amplitudes match those of the largest bipolar electrogram, because the diminishing amplitude effect perpetuated by a perpendicular activation wavefront direction is removed. Yet, omnipolar and bipolar scar maps differ due to the effects of an altered voltage threshold selection, and variations in oncoming wavefront orientation. Although a potentially promising technology, published clinical AF data in which omnipolar mapping overcomes the directional limitation of bipolar voltage mapping is scarce[127]. Since omnipolar electrograms do provide consistent and independent voltage information in AF[125], they may eventually obviate the need for cardioversion to sinus rhythm to undertake voltage mapping.

RHYTHM TYPE

Atrial fibrosis promotes arrhythmogenesis by supporting reentrant and focal mechanisms, and it contributes to the post-ablation transition from AF to more organized rhythms such as atrial flutter and AT[128]. Most post-ablation ATs originate in the LVAs, often coinciding with sites of prior ablation lesioning, and thereby underscore the role of LVAs as a source of conduction block in arrhythmia recurrence[129]. Bipolar voltages are significantly lower in AF than in sinus rhythm or atrial flutter due to the presence of rapid and disorganized activation, shortened cycle lengths, and potential functional refractoriness of the atrial substrate[10,107]. In the LA, voltages during sinus rhythm average around 3.0 ± 2.9 mV, while they drop substantially to approximately 0.9 ± 0.6 mV during AF[130]. Left atrial posterior and right atrial septal regions exhibit the most significant voltage reductions during AF[130]. In many persistent AF patients, LVAs identified in sinus rhythm correspond to those seen in AF, although the extent of LVAs is often larger in AF when using the same threshold voltage for detection[1]. At the posterior LA, AF voltage correlates with fibrotic tissue detected by CMR-GE better than sinus rhythm voltage, likely due to the functional unmasking of latent substrate under the conditions imposed by the arrhythmia[63].

Electrogram fractionation is more prevalent during AF than in sinus rhythm, especially in areas of abnormal substrate. In persistent AF patients, fractionated electrograms occurred in 45% of sites during AF, as compared to 13% in sinus rhythm[30]. Sites exhibiting fractionation during AF were more likely to also show abnormalities during sinus rhythm, suggesting an underlying and fixed substrate at many locations with abnormality. During induced paroxysmal AF patients, shortening of the systolic interval, following either drift or acceleration of a source, results in intermittent fibrillatory conduction with formation of fractionated electrograms at the posterior left atrial wall[131]. Detailed examples of such fractionated AF electrograms are shown in Figures 5, 6, 7 and 8. In each Figure, to the left are the sequential time series traces. To the right are three-dimensional graphs of spectral estimation, showing frequency magnitude (Z-axis), and time vs frequency (X-axes and Y-axes). The upper panel portrays the global view of the entire spectrum of length 16 seconds and 2–12 Hz frequency, while the lower panel depicts a close-up or zoom-in of the frequency range, which varies in each of the Figures 5, 6, 7 and 8 to show detail. Frequency magnitude is color-coded, with low amplitudes blue and highest values red. The DF is often more regular and stable in persistent (Figures 5 and 6) as compared to paroxysmal AF (Figures 7 and 8). This is evident at the peak magnitudes (red, orange and yellow color) for persistent AF, both at anterior and posterior free wall sites (Figures 5 and 6, respectively) vs increased variability in paroxysmal AF (Figures 7 and 8). Thus, there is greater spatiotemporal dispersion in paroxysmal AF spectra. The spatiotemporal dispersion of the electrogram activation sequence reflects reentrant-like conduction, and is indicative of its contribution in initiating or maintaining AF[132].

Figure 5 Left panel-successive time series segments from an atrial fibrillation recording in a persistent atrial fibrillation patient, anterior free wall. The electrogram morphology is highly variable from one cycle to the next, yet the individual cycles are distinguishable, and the cycle length is evident. This observation translates to a similar and approximately constant dominant frequency (DF) spectrum (right panels) which is highlighted as the highest amplitude value (red, orange, and yellow color) over time. The DF for this persistent atrial fibrillation patient is therefore on the order of approximately 5.7 Hz, but it varies and bifurcates.

Figure 6 Left panel: An atrial fibrillation electrogram, posterior free wall, taken from the same persistent atrial fibrillation patient as in Figure 5. Here the signal amplitude is more vacillating (left panel) vs Figure 5, and this translates into a more changeable spectral magnitude (maximum amplitude or color, right panels). As is evident in the lower frequency graph, there is less organization at signal onset (first 8 seconds) with only a low power dominant frequency (green and some yellow color above the blue background). During the latter half of the interval, however, two apparently competing dominant frequencies arise and form large spectral components (red and orange colors).

Figure 7 Paroxysmal atrial fibrillation patient data. To the left is the sequence of successive intervals from a left atrial recording during atrial fibrillation, anterior free wall. The deflections appear to be more randomized (variable deflection shape, duration, and interval between deflections), as compared with those from the persistent AF patient, Figures 5 and 6. Indeed, a more randomized frequency spectrum is evident in the right panels. The dominant frequency (DF) in this case ranges from about 6–6.5 Hz. Thus, it is fluctuating in frequency over time. Furthermore, the DF magnitude or amplitude is quite variable even over short time intervals, particularly evident in the zoom-in lower frequency graph.

Figure 8 In this bipolar signal recording from the same paroxysmal atrial fibrillation patient, posterior free wall, there are longer isoelectric intervals between time series activation deflections (left panel) with irregular morphology. Yet, this substrate may be more organized in the sense of activating more or less simultaneously, though it is less organized in the sense of variability in the deflection morphology and timing of each of these activation events. Overall, it results in the frequency spectrum (right panels) being composed of mostly low-power areas, and very little constancy in the dominant frequency (use of clinical data and its analysis for construction of the Figures 1, 2, 3, 4, 5, 6, 7 and 8 contained herein was approved by the Institutional Review Board of Columbia University Irving Medical Center).

THE P-WAVE

The surface P-wave on a standard 12-lead electrocardiogram (ECG) offers valuable noninvasive insights into atrial conduction and remodeling. Numerous studies have linked P-wave characteristics, including amplitude, duration, axis, and dispersion, to AF onset, recurrence, PR interval, and atrial substrate abnormalities[133,134]. P-wave dispersion (Pd), defined as the difference between the maximum and minimum P-wave durations across 12 ECG leads, reflects the presence of heterogeneity in atrial conduction[135]. Increased Pd measured within 24 hours of an acute ischemic stroke may predict paroxysmal AF and identify patients at risk for recurrent stroke[136]. In stroke patients, Pd normalized to P-wave vector magnitude (Pd/Pvm) is the only ECG parameter independently associated with subsequent AF detection[137]. Lower P-wave voltage in lead 1 (PVL1) corresponds to increased AF recurrence risk following ablation[138]. Furthermore, a PVL1 magnitude less than 0.062 mV predicts the presence of left atrial LVAs overlying more than 35% of the atrial surface[139]. Reduced PVL1, and presence of interatrial block, are both associated with new-onset AF in patients with non-ST elevation myocardial infarction (MI), suggesting a shared pathophysiological mechanism which is linked to the presence of structural heart disease[138]. Prolonged P-wave duration is correlated with increased left atrial activation time, particularly in the presence of low-voltage substrates[140]. Left atrial low-voltage substrate significantly delays regional conduction[141], thus increasing P-wave duration. This contributes to arrhythmia recurrence after PVI, especially in persistent AF. Prolonged P-wave duration and reduced amplitude correspond to the extent of atrial scarring. These parameters may therefore serve as noninvasive markers of electroanatomic remodeling[139,140].

The P-wave duration-to-voltage ratio has emerged as a simple and cost-effective ECG metric. It shows promise in predicting new-onset AF, and may also provide mechanistic insight into atrial conduction abnormality[142]. The ratio integrates conduction time and atrial electrical amplitude, potentially serving as a surrogate for atrial remodeling severity[143]. A lower Pvm has been linked to the presence of extensive LVAs, and it may function as a complementary tool, alongside conventional ECG metrics, in left atrial substrate assessment[139]. An increased left atrial diameter observed on echocardiography is often associated with prolonged P-wave duration, and enhances AF risk stratification[144]. Thus, P-wave characteristics on the standard ECG, including dispersion, duration, voltage, and derived ratios, serve as important noninvasive tools for predicting AF risk, identifying atrial remodeling, and guiding patient selection for ablation[145]. These markers also hold relevance in stroke populations, post-MI patients, and those undergoing rhythm control strategies.

CONCLUSION

Voltage-guided substrate mapping is an important tool for the treatment of AF. Low voltage areas have been detected in a majority of patients with paroxysmal AF[97] and it is also useful for guiding linear ablation in patients with persistent AF[81]. This finding challenges the notion that substrate severity and the need for ablation are solely dependent on the time duration of the arrhythmia since onset. Rather, a median voltage of 0.99 mV in paroxysmal AF patients vs 0.41 mV in patients with persistent AF is useful to inform on AF type[25], thereby suggesting that substrate remodeling is important to AF progression, and in transition between AF types. Further study of remodeling would be useful to determine mechanistically salient aspects of disease progression, and those which are most important for substrate modification. Voltage-guided ablation using the established 0.5 mV cut-point during sinus rhythm for targeting low voltage regions improves clinical outcome, adding credibility to this metric as a useful tool. However, there is as yet no standard cut-point, and regions below the 0.5 mV cutoff in sinus rhythm typically exhibit lower thresholds during AF (approximately 0.2 mV) and atrial flutter (0.35 mV)[10]. Hence, there is a need to develop standards to target for ablation during substrate mapping, and to understand how these are affected by rhythm type. Although a simple peak-to-peak voltage measurement is normally used to determine level, the AF atrial electrogram has a rich morphology which might be considered for further appraisal (Figures 5, 6, 7 and 8). Many facets of atrial structure and of the data acquisition system influence voltage substrate mapping, and these should be considered, and if possible normalized, when comparing and contrasting the results from different studies. Ultimately, personalized ablation of LVA in persistent AF may provide further insight for substrate mapping and modification[146].