Abstract
Free wall locations are the most common positions for accessory pathways (APs) in clinical practice. Right and left free wall APs account for 10% to 20% and 50% to 60% of all APs, respectively. These pathway locations each present distinct challenges to the electrophysiologist. Left free wall APs are amenable to ablation and have the highest success rates and lowest incidences of recurrence. The left heart location is less accessible, however, necessitating arterial or transseptal approaches. In contrast, right free wall APs are readily approached from simple venous access, but have the lowest success rates and highest incidences of recurrence. Newer mapping systems and technology is helping to increase success rate of ablation with less fluoroscopy and fewer complications.
Keywords
ablation, left-sided accessory pathway, right-sided accessory pathway, supraventricular tachycardia, WPW
Key Points
- •
The atrioventricular (AV) annulus is mapped for atrial or ventricular accessory pathway (AP) insertion sites or the AP itself.
- •
Ablation targets include the earliest site of atrial or ventricular activation by the AP, sites of AP potentials, and sites of electrogram polarity reversal (left free wall APs).
- •
Special equipment includes preformed sheaths (especially for the transseptal approach), multielectrode halo mapping catheters, and steerable sheaths (especially for right free wall). Catheter navigation systems are useful, and electroanatomic mapping may play an important role in reduction of fluoroscopy.
- •
Sources of difficulty are misdiagnosis of atrioventricular nodal reentry with eccentric atrial activation or atrial tachycardia as orthodromic reciprocating tachycardia, catheter stability (especially with right free wall APs), AP insertion sites away from the annulus, multiple insertion sites, and epicardial APs.
Free wall locations are the most common positions for accessory pathways (APs) in clinical practice. Right and left free wall APs account for 10% to 20% and 50% to 60% of all APs, respectively. These pathway locations each present distinct challenges to the electrophysiologist. Left free wall APs are very amenable to ablation and have the highest success rates and lowest incidences of recurrence. The left heart location is less accessible, however, necessitating arterial or transseptal approaches. In contrast, right free wall APs are readily approached from simple venous access but have the lowest success rates and highest incidences of recurrence.
Anatomy
The anatomy of the tricuspid annulus is different from that of the mitral annulus. The hinge of the mitral atrioventricular (AV) annulus is a well-formed and distinct cord of fibrous tissue around the annulus ( Fig. 23.1A ). This accretion of fibrous tissue is interposed between the atrial and ventricular myocardia. On the ventricular side of the mitral annulus, basal cords of ventricular myocardium may descend in a curtain-like fashion from the mitral hinge to insert into the trabeculations on the ventricular wall. These cords may limit catheter mobility beneath the valve during attempts at left free wall AP ablation using the retrograde aortic approach. In the limited human histologic descriptions of left free wall APs, the atrial connection is usually discrete and near the annulus. The pathways then skirt the annulus on its epicardial aspect and may cross at variable depths within the epicardial fat pad ( Fig. 23.1B ). The ventricular insertion usually branches into multiple connections with the ventricle that may be displaced away from the annulus, toward the apex. The histologically determined length of an AP is typically 5 to 10 mm, with a maximal diameter of 0.1 to 7 mm. The left-sided epicardial AV groove is shallow but contains the left circumflex artery near the annulus and the coronary sinus (CS) more remote from the annulus. Although the CS is useful for quickly mapping the mitral valve area, it runs an average of 10 to 14 mm on the atrial side of the true annulus. This separation from the annulus is more pronounced in the proximal 20 mm of the CS. Therefore during catheter mapping, the CS location and electrograms are best regarded as gross estimates of the true AP location on the annulus. The anterior limit of the left free wall is anatomically well demarcated by the aorto-mitral valve continuity, which rarely contains AP connections. The exact location of this continuity is difficult to recognize by fluoroscopy alone. The posterior limit of the left free wall is anatomically continuous with the posteroseptal area and is arbitrarily defined on fluoroscopy.
In contrast to the mitral annulus, the tricuspid valve annulus (TVA) is less well formed and frequently discontinuous. The right atrial and right ventricular myocardia tend to overlap or fold over one another as they insert on the tricuspid annulus (see Fig. 23.1 B ). Right free wall APs may cross discontinuities in the less distinct fibrous annulus or skirt the epicardial aspect of the annulus, as do left free wall APs. The less developed tricuspid annulus and acute angulation of the tricuspid leaflets toward the ventricle make catheter positions along the right free wall unstable. Fluoroscopic definition of the right free wall is difficult because there are no clear landmarks for guidance.
Because of the association of Ebstein anomaly with right-sided APs, the anatomy of this condition merits special consideration. In this disorder, the tricuspid valve leaflets are tethered to the ventricular wall for variable distances from the annulus. This contributes to catheter instability during mapping of the tricuspid annulus. Although not anatomically displaced, the true tricuspid annulus may be poorly developed, with extensive discontinuities of the fibrous architecture. Electrograms recorded from the annulus in Ebstein anomaly may be of low amplitude and fragmented owing to the disorganized tissue. This fragmentation adds to the difficulty in mapping the tricuspid annulus in this condition. APs in Ebstein anomaly are often multiple and may skirt the epicardial aspect of the annulus or pass subendocardially through gaps in the fibrous annulus.
A complete description of free wall accessory AV connections must also include those resulting from connections of the ventricle to the CS musculature, the ligament of Marshall, and the atrial appendage. Mahaim-type atriofascicular connections are described in Chapter 26 . The venous wall of the CS is surrounded by a continuous sleeve of atrial myocardium that extends 25 to 51 mm from the CS ostium. This muscle is continuous with the right atrial myocardium proximally but is usually separated from the left atrium by adipose tissue. This separation may be bridged by strands of striated muscle, however, producing electrical continuity between the CS musculature and the left atrium ( Fig. 23.2 ). These connections, which can be broad and very extensive, are reported in up to 80% of hearts in autopsy series. Electrical continuity of the CS musculature with the ventricle is less common, but it may provide the substrate for reciprocating tachycardias. Ventricular connections may result from CS musculature extensions over the middle cardiac vein, posterior cardiac vein, or AV groove branch of the distal left circumflex artery. Ventricular connections with the CS are described in 3% to 6% of hearts at autopsy. Sun and associates reported that 36% of patients (most with previously failed ablation) who had left posterior or left posteroseptal AV connections actually used the CS musculature as the intermediary between the atrium and ventricle in reciprocating tachycardias.
The ligament of Marshall is a vestigial fold of pericardium that carries the vein of Marshall from its origin as a branch of the distal CS to its termination near the left superior pulmonary vein. This structure may also contain bundles of muscle fibers that are continuous with the CS musculature. These fibers may end blindly, or they may insert directly into the left atrial musculature at the inferior interatrial pathway. With proximal connections between the CS musculature and the ventricle, the ligament of Marshall can support AV reciprocating tachycardia.
Another unusual form of AV connection is a direct epicardial muscular continuity between the atrial appendage and the ventricle. Several reports of connections between the right atrial appendage and ventricle are available, whereas left-sided connections appear even more rarely. The developmental basis for these connections is unknown. Because of epicardial ventricular insertions more than 1 cm apical to the annulus and atrial origins within the atrial appendage, endocardial mapping of the annulus for conventional APs is perplexing. For left-sided connections, mapping of the anterior coronary venous branches may demonstrate the earliest ventricular activation. At surgical division of one such right-sided connection, a broad band of myocardium under the epicardial fat pad was noted from the base of the atrial appendage to the base of the right ventricle. Case reports describe successful catheter ablation of these connections from the right atrial appendage.
Pathophysiology
Diagnosis
As with APs at other locations, free wall APs may participate in reciprocating tachycardias or undergo bystander activation during tachycardias mechanistically unrelated to the presence of the AP. Free wall APs have been associated with specific electrophysiologic characteristics. Compared with septal and left free wall locations, right free wall APs are less likely to demonstrate retrograde conduction, to participate in reciprocating tachycardias, and to be associated with inducible atrial fibrillation. Pathways in the right free wall may be more likely to demonstrate decremental anterograde conduction than those in other locations. Compared with right free wall pathways, left free wall APs are more likely to demonstrate decremental retrograde conduction and have longer retrograde refractory periods. Right-sided pathways may worsen cardiac function by activating the RV early and causing a functional LBBB.
Surface electrocardiogram (ECG) localization of manifest free wall APs is imperfect and becomes less accurate if minimal preexcitation is present (QRS <120 ms). ECG algorithms for AP localization are most accurate for the diagnosis of left free wall APs compared with all other locations, achieving at least 90% sensitivity and almost 100% specificity. In using any localization algorithm one must be aware of the portion of the QRS complex on which the algorithm is based. Some algorithms use only the first 20 to 60 ms of the delta wave, whereas others are based on the morphology or polarity of the entire QRS complex. Provided that significant preexcitation is present, all left free wall APs should demonstrate a positive delta wave in V 1 , with the R wave greater than the S wave (R > S) in lead V 1 or V 2 at the latest ( Fig. 23.3 ). A negative delta wave in lead I, aVL, or V 6 is pathognomonic of a left lateral pathway. As the pathway location moves from posterior to lateral to anterior, the delta waves in the inferior leads, especially aVF and III, change from negative to isoelectric to positive in polarity. The ratio of the dominant amplitude in limb lead II and III can predict if a left free wall pathway is anterior lateral or posterior lateral. If the ratio II/III is greater than 1, the pathway is likely anterior, and if it is less than 1, it is likely posterior.
As opposed to left free wall APs, the ECG diagnosis of right free wall APs is the least accurate and least consistent among algorithms, with a sensitivity of 80% to 90% and a specificity of 90% to 100%. Confusion may arise in the interpretation of a positive delta wave in V 1 as indicating a left-sided AP ( Fig. 23.4 ). This finding is diagnostic of a left free wall AP only if R is greater than S; a positive delta wave with R less than S in V 1 is consistent with a right free wall AP or a minimally preexcited left free wall AP. A negative delta wave in V 1 is consistent with a septal AP. Therefore most algorithms identify right free wall APs by a positive initial delta wave in V 1 but a late transition to R greater than S in the precordial leads at V 3 or later, coupled with leftward orientation to the initial delta wave, such as delta wave positivity in lead I or aVL. As the pathway location moves from the right superior free wall to the right middle and right inferior free wall, the delta wave in inferior leads aVF and II changes from positive to isoelectric to negative. A useful algorithm for AP localization that is based on the initial 20 ms of the delta wave is shown in Fig. 23.5 .
A study comparing several algorithms was not able to duplicate the accuracy of the original papers. Wren et al. looked at the accuracy of seven well known algorithms to predict AP location in children. In 100 children ages 3.8 to 17 years (mean age 11.7 years) the accuracy of prediction was 30% to 49% for exact location and 61% to 68% including adjacent locations. The algorithms were least accurate in predicting midseptal or right anteroseptal AP locations. The accuracy of algorithms may vary with ECG lead position, degree of preexcitation, patient body habitus, heart rotation, and prior QRS abnormalities.
The location of the AP can also be inferred from the surface ECG by the polarity of the retrograde P waves during orthodromic reciprocating tachycardia (ORT). A negative P wave in lead I is highly suggestive of a left free wall location, with a 95% positive predictive value. A negative P wave in lead V 1 is highly suggestive of a right-sided AP. The presence of a positive retrograde P wave in lead I suggests a right free wall AP with a positive predictive value of 99%. For either right or left free wall APs, the presence of negative P waves in all three inferior leads indicates an inferior location, whereas positive P waves in these leads indicate a superior location. Isoelectric or biphasic P waves in any of the inferior leads suggest a middle free wall location.
At electrophysiologic testing, the hallmark of any ORT is the demonstration of obligatory 1:1 atrial and ventricular activation for persistence of the tachycardia ( Box 23.1 ). The diagnosis of ORT using a free wall AP also requires an eccentric atrial activation sequence earliest along the right or left atrial free wall. Coupled with such an eccentric atrial activation sequence, ORT is highly suggested by demonstration of the shortest QRS-to-atrium time of 60 ms or more, constant ventricle-to-atrium times despite changes in the tachycardia cycle length, and the ability to advance atrial activation by a premature ventricular stimulus delivered during His bundle refractoriness. The last finding indicates the presence of an AP but does not prove participation in the tachycardia. A preexcitation index greater than 70 ms is consistent with a left lateral AP. The preexcitation index is the difference between the tachycardia cycle length and the longest coupling interval of a right ventricular apical premature stimulus that advances the atrium. Diagnostic of a free wall pathway is prolongation of the QRS-to-atrium (or His-to-atrium) time during reciprocating tachycardia (and usually of the tachycardia cycle length as well) by 35 to 40 ms or longer with ipsilateral bundle branch block. Left anterior fascicular block also prolongs the QRS-to-atrium time in patients with left free wall APs. Coupled with an eccentric atrial activation sequence, the ability to reproducibly terminate the tachycardia with a premature ventricular stimulus delivered during His bundle refractoriness but that does not result in atrial activation also proves an ORT. Parahissian pacing techniques consistently indicate the presence of retrograde conduction over right free wall APs. The response of left free wall APs to parahissian pacing is more complex. In about 25% of cases of left free wall APs, parahissian pacing is consistent with retrograde conduction over only the AV node. His capture may lead to a paradoxical shortening of the ventricle-to-atrium times, with left lateral APs because of the more rapid activation of the left free wall through the His-Purkinje system than occurs with septal ventricular capture alone.
Orthodromic Tachycardia
Obligatory 1:1 AV relationship with earliest atrial activation on AV free wall
Shortest V-to-A time ≥60 ms
Constant V-to-A conduction times despite TCL variations
Advance atrial activation during His refractoriness (proves pathway presence but not participation in tachycardia)
Preexcitation index >70 ms (left lateral AP)
Ipsilateral bundle branch block prolongs His- (or V-) to-A time (and usually TCL) by ≥35 ms a
a Proves free wall AP-mediated tachycardia.
Reproducible tachycardia termination by premature ventricular stimuli during His refractoriness without conduction to atrium b
b Proves AP-mediated tachycardia.
Antidromic Tachycardia
Obligatory 1:1 AV relationship with earliest ventricular activation on free wall
QRS morphology in tachycardia consistent with maximal preexcitation
Tachycardia QRS morphology reproduced by atrial pacing near pathway insertion
Each limb of tachycardia circuit supports conduction at TCL
Advance ventricular activation by atrial extrastimuli near insertion with advancement of subsequent His and atrial activation b
Changes in V-to-His interval precede changes in TCL
Exclusion of ventricular tachycardia and bystander participation, especially AV nodal reentry (His-to-A time in tachycardia ≤70 ms consistent with AV nodal reentry)
A, Atrium; AP, accessory pathway; AV, atrioventricular; His, bundle of His; TCL, tachycardia cycle length; V, ventricle.
The diagnostic features of antidromic reciprocating tachycardia are given in Box 23.1 . Preexcited reciprocating tachycardias may use the AV node or a second AP as the retrograde limb. There are no surface ECG features that are diagnostic of antidromic tachycardia, but the diagnosis is excluded by demonstration of an atrial-to-ventricular relationship other than 1:1.
Differential Diagnosis
ORTs using free wall APs must be differentiated from atrial tachycardias arising from near the AV valve annuli or, rarely, from the CS musculature. Differentiation between atrial tachycardia and ORT is best accomplished by dissociating the ventricles from the tachycardia. The demonstration of a V-A-A-V response after termination of ventricular pacing that entrains the atrium excludes an ORT and confirms an atrial tachycardia. The ability to initiate the tachycardia with ventricular pacing, initiation with a critical AV or ventriculoatrial (VA) interval, and advancement of the same atrial activation sequence with premature ventricular stimuli during His refractoriness are all consistent with an ORT rather than AV nodal reentry.
About 6% of cases of AV nodal reentry are associated with an eccentric atrial activation sequence that is earliest in the posterior or distal CS, with even the shortest VA times being longer than 60 ms ( Fig. 23.6 ). This pattern is easily confused with ORT using a left-sided concealed AP, both at electrophysiologic testing and on the surface ECG, because the retrograde P wave during AV nodal reentrant tachycardia (AVNRT) is negative in leads I and aVL and positive in V 1 . The eccentric atrial activation sequence is usually demonstrated with ventricular pacing as well. The keys to the diagnosis of AVNRT with eccentric atrial activation are demonstration of dual AV nodal physiology, ability to also induce typical AV nodal reentry with concentric atrial activation or variable patterns of retrograde atrial activation in most patients, absence of retrograde VA conduction without isoproterenol, inability to advance the atrium with premature ventricular stimuli during His refractoriness (may require left ventricular pacing), demonstration of only decremental retrograde VA conduction, and ability to dissociate the atrium and ventricle from the tachycardia. Standard slow pathway ablation in the posteroseptal right atrium eliminates the tachycardia in these cases.
The differential diagnosis of an antidromic tachycardia includes ventricular tachycardia and bystander AP participation. Ventricular tachycardia should be diagnosed by the dissociation of the atrium from the tachycardia or a variable His-to-atrium timing relationship without alteration of the tachycardia cycle length. Antidromic tachycardia is diagnosed by demonstrating an obligatory 1:1 atrium-to-ventricle relationship during tachycardia, reproduction of tachycardia QRS morphology by atrial pacing at the presumed AP insertion site, and advancement of the ventricular and subsequent atrial activation by a premature atrial stimulus near the AP site ( Table 23.1 ).
| Study | AP Potential | EGM Stability | Delta-V Interval (ms) | A Amplitude | Local AV Interval (ms) | Local VA Interval (ms) | Best Positive Predictive Value (%) |
|---|---|---|---|---|---|---|---|
| Hindricks et al. 1995 | + | — | Value not specified | — | — | — | 70 |
| Bashir et al. 1993 | + | — | 10 | — | — | — | 20−25 |
| Chen et al. 1992 | + | <10% change in EGM amplitude | 0 | A >1 mV | — | — | 62 |
| Cappato et al. 1994 | + | — | ≤0 | A/V ratio ≥0.1 | ≤40 | — | 87 |
| Xie et al. 1996 | + | — | 0 | — | 30 | ≤30 | 67 |
| Villacastín et al. 1996 | + | — | — | — | — | Pseudo disappearance | 59 |
Bystander participation of the AP is best recognized by dissociation of AP conduction from the tachycardia. The demonstration of a His-to-atrium interval of 70 ms or less indicates AV nodal reentry with bystander AP rather than an antidromic reciprocating tachycardia.
Mapping
The most widely used approaches to mapping of free wall APs rely on identification of the earliest ventricular activation during anterograde AP conduction and earliest retrograde atrial activation during ORT ( Box 23.2 ). However, mapping based on electrogram morphology rather than timing can be performed in some situations. Both unipolar and bipolar recordings are helpful. Unipolar recordings from the electrode tip provide information on local activation through electrogram timing and morphology. Bipolar recordings from the distal electrode pair reflect timing and more clearly demonstrate the electrogram components and AP potentials. Three-dimensional computer mapping has become the standard imaging tool for mapping APs and for marking the best sites to return to if catheter stability is poor or pathway conduction returns shortly after ablation has been transiently successful.
Left Free Wall
Presumed AP potential
Delta-VEGM ≤0 ms (anterograde conduction)
VEGM–AEGM ≤40 ms (retrograde conduction)
AEGM–VEGM ≤40 ms (anterograde conduction)
AEGM amplitude >0.4 mV
QRS–AEGM interval ≤70 ms (retrograde conduction)
Isoelectric VEGM–AEGM interval ≤5 ms (retrograde conduction)
Site of AEGM polarity reversal (in tachycardia)
Right Free Wall
Presumed AP potential
Delta-VEGM ≤ −10 ms
AEGM amplitude >1 mV
AEGM–VEGM ≤40 ms (anterograde conduction)
QRS–AEGM interval ≤70 ms (retrograde conduction)
Isoelectric VEGM–AEGM interval ≤5 ms (retrograde conduction)
AEGM, Atrial electrogram; AP, accessory pathway; Delta, delta wave onset; QRS, QRS onset; VEGM, ventricular electrogram.
Left Free Wall Accessory Pathways
Mapping of left free wall APs is facilitated by multielectrode recordings from a CS catheter; however, the anatomic distance from the true mitral annulus limits the accuracy of CS mapping alone to identify target sites for ablation ( Fig. 23.7 ). Left free wall APs can be ablated without the use of CS catheters (single catheter technique); however, the CS catheter is standard, along with right atrial, His, and right ventricular catheters. Mapping and ablation may be performed by the transaortic (retrograde) approach or the transseptal approach. The transaortic approach is directed at sites beneath the mitral annulus and therefore targets the AP ventricular insertion. For the transaortic approach, the catheter is always prolapsed across the aortic valve to prevent perforation of the leaflets or entry into the coronary arteries. The catheter most readily crosses the aortic valve, with the D curvature of the deflected catheter tip opening to the right of the fluoroscopy screen (anteriorly) in the right anterior oblique view. After entering the left ventricular cavity, the curvature is maintained on the catheter tip, and the catheter is rotated in a counterclockwise direction to turn the catheter tip posteriorly toward the annulus ( Fig. 23.8 and 
on Expert Consult). The catheter then can be opened slightly to engage a subannular position, or it can be withdrawn to cross over the annulus into the left atrium. The catheter tip is either moved incrementally in steps beneath the annulus or made to slide along the mitral annulus for mapping before being dropped down beneath the annulus for energy delivery (see Fig. 23.8 ). It is often difficult to achieve stable catheter positions for the far lateral and anterior mitral annulus with the transaortic approach.
The transseptal approach is primarily directed at mapping the atrial side of the annulus or the mitral annulus itself. As opposed to the transaortic approach, in which the ablation electrode is perpendicular to and beneath the annulus, the transseptal approach is directed at positions on or above the annulus, with the electrodes parallel to the annulus ( Fig. 23.9 ). After passing through the atrial septum, the catheter is directed laterally with a large sweeping curve to direct the tip back toward the atrial septum. With the aid of preformed sheaths, the catheter tip is made to slide along the annulus by advancing and withdrawing the catheter (see Fig. 23.9 and 
on Expert Consult). Anterior mitral valve positions are readily reached with the transseptal approach ( Fig. 23.10 ). Although ability to maneuver the catheter is greater, catheter stability may be more problematic than with the transaortic approach.
The decision regarding use of the transaortic or transseptal approach is based on physician familiarity and certain patient characteristics. Because the two approaches are complementary, it is best for the physician to be familiar with both techniques. The transseptal approach is favored in the presence of peripheral vascular disease, aortic valve disease or prosthesis, or small ventricular chambers. In children weighing less than 30 kg, the transaortic approach may be associated with frequent valve trauma. The transseptal approach provides better access to far lateral and anterolateral AP locations. The transseptal approach may be contraindicated in the presence of distorted cardiac anatomy such as congenital heart disease, pneumonectomy, kyphoscoliosis, or severe dilation of the aorta or right atrium. Intracardiac echo or transesophageal echocardiogram (TEE) would be necessary for these cases.
The characteristics of electrograms at successful ablation sites have been studied extensively for left free wall APs. Most of these data are derived from early experience with ablation by the transaortic approach and have not been reproduced for the transseptal approach. Five electrogram characteristics have been described as useful in predicting successful ablation sites when mapping anterograde AP activation by the transaortic approach: (1) delta-to-ventricle (V) interval, (2) atrial electrogram amplitude, (3) electrogram stability, (4) local AV electrogram interval, and (5) presence of a presumed AP potential. The delta-to-V interval should be measured from the onset of the delta wave to the peak or intrinsicoid deflection of bipolar mapping electrograms. For unipolar electrograms, the maximal negative dV/dt reflects local ventricular activation. The absence of a QS morphology in the unipolar electrogram indicates a site with a less than 10% chance of ablation success. Pacing from near the AP insertion accentuates the degree of preexcitation. For left free wall APs, a delta-to-V time of 0 ms or less is usually recorded at successful ablation sites, with average intervals of only 2 to 10 ms ( Fig. 23.11 ).
The atrial electrogram amplitude at successful sites should be greater than 0.4 to 1 mV, or the atrial-to-ventricular (A/V) ratio should be greater than 0.1 by the transaortic approach. The small atrial electrogram amplitude indicates the subannular position of the catheter. The absence of any atrial electrogram suggests a position too far from the annulus. For transseptal mapping, an A/V ratio equal to 1 is sought along the mitral annulus. Greater and lesser ratios indicate atrial and ventricular displacement of the catheter, respectively. Electrogram stability is defined as less than 10% change in the atrial and ventricular electrogram amplitudes or A/V ratio over 5 to 10 cardiac cycles.
The transaortic approach usually provides good catheter stability because the catheter is “wedged” beneath the mitral valve. For the transseptal approach, stability can be assessed by consistent electrogram amplitudes, catheter motion concordant with the CS catheter, and when on the atrial side of the annulus, consistent PR segment elevation on the unipolar electrogram. The last finding also indicates sufficient catheter contact with the atrial myocardium. Local AV intervals of 40 ms or less should be sought, with average intervals of 25 to 50 ms reported for successful ablation sites (see Fig. 23.11 ). The AV intervals for left free wall APs are longer than for other AP locations, and the times at successful sites overlap with those at unsuccessful sites. The local AV times alone therefore have limited specificity in identifying successful ablation sites for left-sided APs.
Finally, the recording of a presumed AP potential has been used to define appropriate ablation sites. The ability to record AP potentials may be influenced by the catheter approach used to map the annulus. Presumed AP potentials are defined as discrete, high-frequency potentials that occur between the atrial and ventricular electrograms and at least 10 ms before the onset of the delta wave (see Fig. 23.11 ). The amplitude of AP potentials ranges from 0.5 to 1 mV. The validation of a signal as a true AP potential is a tedious process and is not always practical in the clinical setting. It is probably for this reason that AP potentials are reported at 35% to 94% of successful ablation sites and at up to 72% of unsuccessful sites. The recording of larger AP potentials in the CS than along the mitral annulus suggests the presence of an epicardial AP.
For mapping of retrograde AP conduction during ORT or ventricular pacing, electrogram characteristics reported to identify successful sites include catheter stability, presence of a presumed AP potential, continuous electrical activity, and the local VA interval. These data are derived largely from studies of ablation from the transaortic approach. Catheter stability is defined the same as for mapping of anterograde conduction. The presence of a presumed AP potential is reported at only 37% to 67% of successful ablation sites during retrograde AP conduction mapping. It is possible that AP potentials are obscured by the large ventricular electrogram with the transaortic approach. The QRS onset-to-local atrial electrogram (QRS-A) interval is usually about 70 ms in the absence of a left ventricular conduction delay ( Fig. 23.12 ). The local VA interval at successful ablation sites is typically 25 to 50 ms. At very short VA intervals, the atrial electrogram may be inscribed on the terminal portion of the ventricular electrogram. Continuous electrical activity (<5 ms isoelectric interval between V and A electrograms) and the pseudo-disappearance of the atrial electrogram into the ventricular electrogram are manifestations of extremely short VA times. One group has reported successful ablation by pace mapping beneath the mitral annulus with the ablation catheter and measuring the stimulus-to-atrial times recorded on the CS catheter. The site producing the shortest stimulus-to-atrial interval should be at the ventricular insertion of the AP. The average stimulus-to-atrial interval at successful sites was 46 ± 15 ms.
The predictive accuracy of any single electrogram characteristic for identifying a successful ablation site rarely exceeds 30%. The electrogram must satisfy three or four criteria to achieve 60% to 80% predictive values. Multivariate predictors of successful ablation sites are given in Table 23.1 .
One problem with mapping retrograde atrial activation times is the difficulty in discriminating between atrial and ventricular components of the electrograms. This problem may be addressed in several ways. One method is to compare the electrogram recorded on the ablation catheter while simultaneously pacing the atrium and the ventricle with the electrogram recorded during ventricular pacing alone ( Fig. 23.13 ).
