REVIEW Diagnostic Pacing Maneuvers for Supraventricular Tachycardias: Part 2 GEORGE D. VEENHUYZEN, M.D., F. RUSSELL QUINN, M.R.C.P., PH.D., STEPHEN B. WILTON, M.D., ROBIN CLEGG, M.D., and L. BRENT MITCHELL, M.D. From the Libin Cardiovascular Institute of Alberta, University of Calgary and Calgary Health Region, Alberta, Canada The approach to supraventricular tachycardia (SVT) diagnosis can be complex because it involves synthesizing baseline electrophysiologic features, features of the SVT, and the response(s) to pacing maneuvers. In this two-part review, we will mainly explore the latter while recognizing that neither of the former can be ignored, for they provide the context in which diagnostic pacing maneuvers must be correctly chosen and interpreted. Part 1 involved a detailed consideration of ventricular overdrive pacing, since this pacing maneuver provides the diagnosis in the majority of cases. In Part 2, other diagnostic pacing maneuvers that might be helpful when ventricular overdrive pacing is not diagnostic or appropriate, including attempts to reset SVT with single atrial or ventricular beats, para-Hisian pacing, apex versus base pacing, and atrial overdrive pacing, are discussed, as are some specific diagnostic SVT challenges encountered in the electrophysiology lab. There is considerable literature on this topic, and this review is by no means meant to be all-encompassing. Rather, we hope to clearly explain and illustrate the physiology, strengths, and weaknesses of what we consider to be the most important and commonly employed diagnostic pacing maneuvers, that is, those that trainees in cardiac electrophysiology should be well familiar with at a minimum. (PACE 2012; 35:757–769) ablation, electrophysiology - clinical, SVT, pacing In part 1 of this review on diagnostic pacing maneuvers for supraventricular tachycardia (SVT), we explored ventricular overdrive pacing (VOP) in detail, since it provides a firm SVT diagnosis in the majority of cases.1 We will now consider pacing maneuvers that can be performed when VOP is not diagnostic, including ones that can be performed when sustained, regular SVT cannot be induced. These will include singlepaced ventricular beats during ongoing SVT, paraHisian pacing, and apex versus base pacing. We will also explore some challenging specific situations in SVT diagnosis including differentiating atrioventricular node reentry tachycardia (AVNRT) from atrial tachycardia (AT) and junctional tachycardia (JT), SVT with atrioventricular (AV) dissociation, and differentiating AVNRT with a leftward atrionodal exit from orthodromic atrioventricular reciprocating tachycardia (AVRT) employing a left-sided accessory pathway (AP). Address for reprints: George D. Veenhuyzen, M.D., F.R.C.P.C., Libin Cardiovascular Institute of Alberta, University of Calgary and Calgary Health Region, Foothills Medical Centre, Rm C836, 1403-29 St. N.W., Calgary, Alberta, T2N 2T9, Canada. Fax: 403944-1592; e-mail: george.veenhuyzen@calgaryhealthregion.ca Received September 19, 2011; revised December 22, 2011; accepted January 5, 2012. Scanning diastole with ventricular premature beats (VPBs) Single VPB introduced decrementally during diastole in SVT offer an opportunity to determine the relationship between altered timing of ventricular depolarization and the timing of atrial depolarization. For example, if a VPB is able to terminate tachycardia without atrial depolarization, then AT can be excluded, provided this is not a coincidence. Furthermore, VPBs that occur during SVT at a time when the stimulated wavefront would be expected to collide with the SVT wavefront in the His-Purkinje network or in ventricular myocardium cannot possibly affect atrial timing during either AVNRT or AT (unless a bystander AP is present). Accordingly, such Hisrefractory VPBs (HRVPBs) should only be capable of affecting AVRT circuits (again, in the absence of a bystander AP). VPBs that occur before His bundle refractoriness are potentially capable of affecting atrial timing (including terminating SVT) in any of AT, AVNRT, or AVRT. How does one determine if a paced VPB is His-refractory? If the QRS complex morphology of the VPB shows evidence of fusion (i.e., the QRS complex morphology of the VPB shows some features of the QRS complex morphology of a paced VPB and some features of the QRS complex morphology of the SVT), then the paced VPB must be His-refractory, since the SVT wavefront that the doi: 10.1111/j.1540-8159.2012.03352.x  C 2012 Wiley Periodicals, Inc. C 2012, The Authors. Journal compilation  PACE, Vol. 35 June 2012 757 VEENHUYZEN, ET AL. earlier than the anticipated timing of the antegrade His potential, it would not be considered to be His-refractory, but rather, prior to His bundle refractoriness. There are three responses to an HRVPB that are diagnostically useful: Figure 1. Advancement of atrial activation by a fused His-refractory ventricular premature beat (HRVPB). Panel A: During supraventricular tachycardia with a stable cycle length of 444 ms, one-to-one atrioventricular relationship, and an earlier atrial electrogram recorded in the right atrium (where the ablation catheter, ABLp/d, is located) than in the septum (d/pHIS) or coronary sinus (proximal CS 9,10 through distal CS 1,2), a paced premature beat is delivered by electrodes at the right ventricular apex (RVd). The subsequent atrial activation is advanced by 19 ms. The paced premature ventricular beat is His-refractory (HRVPB) because (1) it is fused: note QRS complex morphology features and duration (122 ms) that are intermediate between those of the conducted SVT (narrow complex) and of a purely paced QRS complex (Panel B, QRS complex duration = 144 ms) and (2) the pacing stimulus is delivered at precisely the time of the expected His bundle potential (arrow). A ventricular paced beat can be considered His-refractory if it occurs up to 35–55 ms earlier than the anticipated His bundle potential. Advancement of atrial activation without a change in the atrial activation sequence by an HRVPB indicates that an accessory pathway (AP) is present and almost certainly participating in orthodromic AVRT, in this case, employing a right-sided AP. stimulated wavefront is fusing with in ventricular myocardium must have exited the His-Purkinje network (Fig. 1). If the pacing stimulus occurs just after a discernible antegrade His potential, then the paced VPB is obviously His-refractory. Finally, if the paced VPB occurs no more than 35–55 ms earlier than the anticipated timing of the antegrade His potential, in the time that would be required for that stimulated wavefront to enter the distal arborization of the His-Purkinje network and travel retrogradely to the His bundle, the tachycardia wavefront would have reached the His bundle where these wavefronts would collide.2 When the paced VPB occurs more than 35–55 ms 758 (1) SVT terminates without conduction to the atrium. This response indicates a diagnosis of AVRT, provided that this event is not a coincidence, as could be the case if the SVT frequently spontaneously terminates. The fact that an HRVPB can affect the SVT indicates that an AP is present, and the fact that the SVT terminates with ventriculoatrial (VA) block indicates that ventricular and atrial activation must be linked so that if conduction to the atrium does not occur via the AP, the circuit is interrupted. This cannot be the case with AT or AVNRT, even if a bystander AV AP is present. That is, the AP must also be participating in the SVT mechanism. Theoretically, this response could be observed given the coexistence of AVNRT and a bystander nodoventricular AP, but this occurrence has not been convincingly demonstrated to our knowledge, and would have to be extremely rare. As we discussed in Part 1, sometimes VOP results in an apparently noninterpretable response when VOP repeatedly terminates the SVT. VOP may be considered as a series of consecutive VPBs. When the SVT repeatedly stops during VOP because of VA block, if the paced beat that precedes VA block is His-refractory, and the atrial timing has not changed prior to that VPB, this constitutes an equivalent of an HRVPB terminating SVT without conduction to the atrium, thereby establishing a diagnosis of AVRT (see “What if the response to VOP is not interpretable?” in Part I). (2) Atrial activation is delayed without a change in the atrial activation sequence. This response indicates a diagnosis of AVRT employing a decremental AP (Fig. 2). As above, the fact that an HRVPB can affect the SVT indicates that an AP is present and delay of atrial timing indicates that atrial activation is decrementally linked to ventricular activation. This cannot be the case with either AT or AVNRT even if a bystander AV AP is present; the decremental AP must also be participating in the SVT mechanism. Note that for this response to be appreciated, the degree of decremental conduction slowing must exceed the prematurity of the HRVPB; if they are matched, AVRT employing a decremental AP could be present but because no change in atrial timing would occur, one would conclude that the VPB had no effect on the SVT and the diagnosis could be missed. Accordingly, it is worth studying the effects of multiple HRVPBs June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 Figure 2. Delay of atrial activation by a His-refractory ventricular premature beat (HRVPB). During a long R-P interval supraventricular tachycardia with a stable cycle length of 473 ms, a paced premature beat is delivered by electrodes at the right ventricle. The subsequent atrial activation is delayed by 20 ms without a change in the atrial activation sequence. The paced premature ventricular beat is His-refractory because it occurs virtually simultaneously with and certainly not more than 35–55 ms earlier than the expected inscription of the anterograde His bundle potential (arrow). This response indicates that an accessory pathway (AP) is present and participating in orthodromic AVRT, in this case, employing a slowly conducting concealed septal AP. (Tracing courtesy of Dr. G. Neal Kay.) introduced throughout the His-refractory diastolic window to minimize this potential pitfall of studying only one HRVPB that had no apparent effect. Theoretically, delayed atrial timing after an HRVPB could be observed in the setting of AVNRT with a bystander nodoventricular AP.3 This is so rare that delay of atrial timing by an HRVPB without a change in the atrial activation sequence should be considered extremely strong evidence that the SVT mechanism is AVRT employing a decremental AP. It is often the case that AVRT employing a decremental AP manifests as a long RP interval SVT. If the AP used for retrograde conduction has decremental conduction properties, entrainment by VOP could be associated with long corrected postpacing interval-tachycardia cycle length (cPPI-TCL) and stimulus-atrial (SA)-VA interval values that would normally be considered evidence of atypical AVNRT. Fusion during entrainment would still provide proof that the mechanism is AVRT but, if fusion is not present, it is important to scan diastole with VPBs during PACE, Vol. 35 long RP interval SVTs, even if entrainment by VOP is associated with long cPPI-TCL and SAVA interval values. In such a situation, the finding that an HRVPB delays atrial timing (indicating a diagnosis of AVRT, Fig. 2) has a greater diagnostic value than the finding of long cPPI-TCL or SA-VA interval values (suggesting a diagnosis of AVNRT). Application of this maneuver requires that any apparent delay in atrial timing exceeds the spontaneous variability in the SVT cycle length (CL). Accordingly, this maneuver may not be reliable in irregular SVTs. (3) Atrial activation is advanced without a change in the atrial activation sequence. The fact that an HRVPB can affect atrial timing indicates that an AP is present. If the atrial activation sequence is unaltered, one can conclude with confidence, but not with certainty, that the AP is participating in the SVT mechanism, establishing a diagnosis of AVRT (Fig. 1). Theoretically, AT or AVNRT could be advanced by conduction over a bystander AP, and if the bystander AP were close to the AT origin or atrionodal exit, respectively, the atrial activation sequence may not change appreciably. This situation is so rare that this finding is considered very strong evidence (but not proof) that the SVT mechanism is AVRT. As before, application of this maneuver requires that any apparent change in atrial timing exceeds the degree of spontaneous variability in the SVT CL. Accordingly, this maneuver may not be reliable in irregular SVTs. Unfortunately, while these responses to HRVPBs are specific (or, in the third case, nearly specific) for AVRT, they are not particularly sensitive. If the pacing site is far from the participating AP, the orthodromic wavefront of the VPB may not have had enough time to reach the AP and affect the AVRT circuit when delivered late enough to be His-refractory. The classic situation in which an HRVPB delivered from the right ventricular (RV) apex does not affect an AVRT circuit because of the distance of the RV apex to the AP occurs when a left free wall AP is operative; nevertheless, this problem may arise when a relatively nearby septal AP is involved.4 As is the case with fusion during VOP, the sensitivity of the three responses to HRVPBs described earlier can be increased by moving the pacing site close to the AP, that is, to a basal ventricular site close to the site of earliest atrial activation. This is not surprising, since fusion during VOP constitutes the continuous resetting of an AVRT circuit by a series of consecutive HRVPBs; the fact that they are fused proves that they are His-refractory, as discussed earlier. June 2012 759 VEENHUYZEN, ET AL. It is commonly taught that if none of the three responses to HRVPBs described earlier is observed, no conclusion regarding the SVT mechanism can be drawn. However, it has been suggested that, in the case of a short RP interval SVT, advancing the local ventricular activation adjacent to the earliest atrial activation by more than 30 ms without affecting atrial timing should exclude the participation of a conventional AP from the SVT mechanism.5 Similarly, in the case of a long RP interval SVT, advancing the local ventricular activation adjacent to the earliest atrial activation by more than 60 ms without affecting atrial timing should exclude the participation of a decremental AP from the SVT mechanism.5 Regarding the HRVPB maneuver, one point seems to be underappreciated: if an SVT that resembles typical AVNRT is induced (central atrial activation and septal VA < 70 ms), this maneuver will not add further diagnostic information as it cannot distinguish AVNRT from AT (see one exception for this in the section on AV-dissociated SVT). If the diagnosis is not clear from findings during spontaneous perturbations in the SVT, the pacing maneuver of choice in this situation is VOP. Para-Hisian & Pure-Hisian Pacing The pacing maneuvers described earlier rely on studying the response to an induced perturbation (VOP or single premature beats) of a stable tachycardia. However, it is common to encounter SVTs that are difficult to induce, nonsustained, irregular, or repeatedly terminate during pacing protocols. Such circumstances may prevent diagnostic pacing maneuvers from being performed, or can limit interpretation of their results. In a patient with documented SVT, but no preexcitation on their baseline ECG, one of the goals at an electrophysiologic study is to determine the presence or absence of a concealed AP. Sometimes programmed stimulation from the RV apex (or even catheter-induced PVCs) can quickly give a clue: if the atrial activation sequence is clearly “eccentric” (either right or left free wall), then an AP is very likely to be present. Further characterization of the conduction properties will be required to be more certain, but with little effort, something of interest will have been discovered and further investigations can be directed accordingly. A CS catheter is commonly placed at the start of an electrophysiology study, so this works well for detecting nonseptal left-sided APs, which account for around 50% of all APs.6 The remainder of APs can be harder to detect by this method. With “standard” catheter positions there is not usually a catheter recording 760 from sites all around the tricuspid annulus, so both septal and right free wall APs, as well as retrograde AV nodal conduction, can show earliest atrial activation on the His catheter or at the proximal CS. Another pacing maneuver—paraHisian pacing—can be useful in these circumstances.7 The basic concept is simple—pacing is performed next to the His bundle/proximal right bundle (HB-RB) and the response is studied when the HB-RB and adjacent myocardium are captured, versus when local myocardium alone is captured. This is usually achieved by varying the pacing output and examining the surface QRS duration and, where possible, the stimulusHis (SH) interval. HB-RB capture will produce a narrow QRS complex and short SH interval, whereas loss of HB-RB capture will produce a wider QRS complex and lengthening of the SH interval. In the latter circumstance, the His bundle will only be activated after excitation has traveled through ventricular myocardium, penetrated the distal Purkinje network, and traveled retrogradely through the conduction system. Often, with HBRB capture, the His potential is difficult to discern (particularly when a single catheter is used for pacing and recording), either due to saturation of the His channel by the pacing stimulus, or masking by the local ventricular potential. Appearance of a clear retrograde His potential, however, is usually an indication that HB-RB capture has been lost. We can now consider the response (timing and pattern of retrograde atrial activation) in the absence and presence of an AP (Fig. 3). If AV nodal conduction alone is present then loss of HB-RB capture will cause a lengthening of the stimulusatrial (SA) interval (because excitation has a longer path to travel back to the atrium), without a change in the atrial activation sequence. The change in SA interval should match the change in the SH interval, when this can be measured, and the His-atrial (HA) interval should be the same. If AP conduction alone is responsible for retrograde atrial activation (i.e. no VA conduction is present through the AV node), then loss of HB-RB capture should have little or no effect on the SA interval and no change in the pattern of atrial activation. The SH interval will still lengthen when HB-RB capture is lost, thus the HA interval will shorten, since atrial timing depends only on conduction over the AP. When both AV nodal and AP conduction is present then a more complex response may be obtained, depending on the proximity of the AP to the pacing site and the conduction properties of the AP and the AV node/conduction system. Generally, a change in the retrograde atrial activation sequence should be seen, although this will depend on how much of the atrium is activated via the AP versus the June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 proximal poles of the CS catheter but some delay in atrial timing on the His catheter. Pitfalls Figure 3. Responses to para-Hisian pacing. In both panels, pacing is being performed from the distal poles of the His catheter (dHis). The first beat in each panel captures the His bundle and local ventricular myocardium (narrow QRS complex), whereas the second beat loses His capture and only stimulates ventricular myocardium. Panel A shows the response when retrograde conduction is occurring over a concealed accessory pathway; with loss of His capture there is no change in the SA interval (the time from stimulus [dotted line] to earliest atrial activation [dashed line]), nor is there a change in the atrial activation sequence. In this case, a right para-Hisian pathway was present, with earliest atrial activation on the HRA catheter. Panel B shows the response after successful ablation of the accessory pathway, demonstrating the response when purely AV nodal retrograde conduction is present. With loss of His capture the SA interval extends by 61 ms since the stimulated wavefront must now travel through ventricular myocardium, penetrate the distal branches of the His-Purkinje system, then travel retrogradely through the AV node. Earliest atrial activation is tied between the proximal bipole of the His catheter (pHis) and the CS os (CS 9–10). HRA = high right atrium; CS = coronary sinus; RVA = right ventricular apex; QRSd = QRS complex duration (ms); SA = time from stimulus to atrial electrogram (ms). AV node, relative to the position of the recording sites in the atrium (i.e. where fusion is occurring in the atrium in each case). Demonstrating this change is facilitated by having a catheter close to the site of earliest retrograde atrial activation (i.e. some additional mapping in the atrium may be required). If the AP is close enough to the pacing site and has sufficiently rapid conduction then the SA interval should remain the same with loss of HB-RB capture, but there will be a change in the atrial activation sequence. For example, if a posteroseptal AP is present then loss of HBRB capture may lead to a similar SA interval on PACE, Vol. 35 (1) APs distant from the pacing site: Interpretation of the response to para-Hisian pacing has been shown to be reliable for septal and right free wall APs, but can be misleading for left lateral APs. In the latter case, the pathway may be so far from the pacing site that the atria are entirely activated via the AV node whether the HB-RB is captured or not (assuming AV nodal conduction is sufficiently rapid). However, as previously mentioned, an eccentric atrial activation sequence may be clearly apparent for left-sided APs simply with RV apical pacing or programmed stimulation. (2) Slowly-conducting APs: Similarly, if conduction over an AP is slow relative to AV nodal conduction then the response to para-Hisian pacing may falsely suggest AV nodal conduction alone.7,8 (3) Lack of ventricular capture during HBRB pacing: Occasionally, pure-Hisian pacing can occur, without capture of the local ventricular myocardium (sometimes called “reverse paraHisian pacing”). This phenomenon is usually transient, but can be associated with changes in the QRS duration and if not recognized can lead to a misinterpretation of the response. If it is recognized, then the response can be analyzed and can also give diagnostic information.9 (4) Presence of a fasciculoventricular connection: These rare pathways connect the proximal conduction system to basal septal myocardium and, if present, can prevent low output pacing from capturing myocardium alone; even with loss of direct His bundle capture, excitation can still reach the conduction system so little change in QRS duration may be seen.10 (5) Loss of capture of the proximal left bundle branch alone: This can cause QRS widening without loss of retrograde conduction to the AV node and if not recognized could lead to misinterpretation of the response.11 (6) Inadvertent atrial capture: This can give the impression that retrograde conduction is via the AV node when a septal AP is present, and it can also give the impression that retrograde conduction is via a septal AP when no such AP is present. Atrial capture is best identified by noting a change in atrial timing when adjusting the catheter basally (to deliberately capture the atrium and reduce the interval from the pacing stimulus to a septal atrial electrogram) or apically (to deliberately lose capture of the atrium and prolong the interval from the pacing stimulus to a septal atrial electrogram by more than June 2012 761 VEENHUYZEN, ET AL. 20 ms).12 Aditionally, it seems that when the interval from the pacing stimulus to the atrial electrogram recorded by the proximal CS electrodes is < 60 ms (or < 70 ms to the atrial electrogram recorded by the high right atrial electrodes), atrial capture is almost certainly present, and when these intervals exceed 90 and 100 ms, respectively, atrial capture is almost certainly not present.12 Another fundamental issue with para-Hisian pacing is that demonstrating the presence of an AP does not prove that it participates in SVT— the maneuver alone cannot distinguish between an AP that participates in AVRT from one which is a bystander. However, in the presence of a stable sustained tachycardia, entrainment pacing can also be performed from the para-Hisian region and the response can demonstrate whether an AP is part of the SVT circuit (para-Hisian entrainment), though the details remain beyond the scope of this review.13 Apex versus Base Pacing Another pacing maneuver that can help to disclose the presence of a retrogradely conducting septal AP even in the absence of inducible sustained SVT is apex versus posterobasal pacing.14 This maneuver is, in our opinion, easier to perform and easier to interpret than para-Hisian pacing. Just like para-Hisian pacing, apex versus posterobasal pacing takes advantage of the shorter VA conduction time expected with VA conduction via a septal AP during basal pacing than apical pacing. This maneuver was first described by Martinez-Alday in an elegant study where apical and right posterobasal pacing were performed either in sinus rhythm or as VOP during SVT in patients with posteroseptal APs (resulting in entrainment with fusion in the majority of cases) and patients without posteroseptal APs. Care was taken to avoid atrial capture during pacing at the right posterobasal site.14 The VA index (VAI) was defined as the VA interval (measured from the pacing stimulus artifact to a stable reference at the high right atrium) after pacing from the RV apex minus the VA interval after pacing from the RV base (Fig. 4). All patients with a septal AP had a positive VAI. While most patients without septal APs had negative VAIs, surprisingly, a couple of patients without septal APs had VAIs of 0 or +5 ms, most likely indicating some anatomic and/or physiologic heterogeneity in the retrograde input of the His-Purkine network. Nevertheless, a VAI > 10 ms had 100% sensitivity and specificity for a septal AP in that small study. This discriminatory value is close to the VAI of 0–5 ms observed in two patients without septal APs, so care should 762 Figure 4. Apex versus base pacing consistent with the presence of an accessory pathway. Panel A shows right ventricular (RV) apical pacing (note the superior QRS complex frontal plane axis) with a ventriculoatrial (VA) interval of 142 ms measured from the pacing stimulus to the earliest atrial electrogram recorded by electrodes along the middle of the coronary sinus catheter (CS1,2 = distal; CS 9,10 = proximal). Panel B shows pacing from the basal RV (note the inferior QRS complex frontal plane axis) with a resulting VA interval of 120 ms. The difference between these values is the VA Index (+22 ms), which is consistent with VA conduction over an accessory pathway. probably be taken in reaching firm conclusions based on borderline VAI values. Like para-Hisian pacing, this maneuver can be limited in the detection of a slowly conducting AP. Also, because retrograde conduction can fuse over an AP and the normal AV conduction system, this maneuver should probably be limited to the identification of posteroseptal APs. Because differential entrainment (discussed in Part 1 of this Review) employs the same retrograde pathway as the tachycardia, this limitation may not apply, and differential entrainment certainly did appear to be diagnostically useful in patients with nonseptal APs.15 AVNRT versus AT Distinguishing AVNRT from AT is usually problematic when VOP does not accelerate the atria to the pacing CL (the ventricles are dissociated from the atria) and the SVT has a 1:1 AV relationship with central atrial activation. The ability to dissociate the ventricles from the SVT mechanism excludes the participation of an AP in the SVT mechanism, but AVNRT must still be June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 distinguished from a septal AT. It is noteworthy that in approximately 80% of such cases, the diagnosis is AT,16 but the AT mechanism is by no means proven. Strictly speaking, the VA interval cannot differentiate septal AT from AVNRT as any VA interval is possible with either mechanism. Nevertheless, virtually simultaneous atrial and ventricular activation, as is observed in typical AVNRT, has a very high positive predictive value for that diagnosis, largely because AVNRT is so much more common than AT (predictive accuracy of a test is influenced by prevalence).16 Nevertheless, it is possible for the AV relationship in a septal AT to coincidentally mimic that of typical AVNRT.17 Other SVT features that may be useful to distinguish AVNRT from AT include: (1) The termination (either spontaneously, or after a vagal maneuver or adenosine administration) of SVT on a nonpremature atrial electrogram implies that termination is associated with AV block. This observation favors AVNRT but could occur by coincidence in AT. (2) Termination of SVT on a nonpremature atrial electrogram by ice-mapping in the region of the slow AVN pathway implies that termination is associated with AV block and also favors AVNRT.18 (3) Continuation of the SVT despite AV block favors AT but AVNRT with AV block (usually infranodal) can occur. (4) Termination of SVT after a VPB that does not conduct to the atrium favors AVNRT but could occur by coincidence in AT. (5) When there are small variations in TCL, if HH or VV interval changes precede and predict AA interval changes (i.e. the HA or VA interval is constant despite HH or VV interval changes), then a diagnosis of AVNRT can be made. (6) The apparent requirement of SVT induction upon a “jump” to the AVN slow pathway favors AVNRT but does not prove this diagnosis. (7) An AV Wenckebach CL that exceeds the tachycardia CL favors AVNRT. In addition to the features described earlier, atrial overdrive pacing (AOP) may be useful in this situation (Fig. 5). If, after AOP at a CL 10–40 ms shorter than the SVT CL, the VA interval on the first return beat of the entrained SVT is within 10 ms of the VA interval of the SVT (“VA linking”), a diagnosis of AVNRT is favored. Linking of atrial and ventricular activation would not be expected in AT.16 Unfortunately, it is not uncommon for the VA interval on the first return beat of AVNRT to vary by more than 10 ms (just as it can in the first few beats after the induction of AVNRT).16 Moreover, on rare occasions, the PACE, Vol. 35 appearance of VA linking could occur during an AT by coincidence.16 Accordingly, this finding should be considered strong evidence, rather than proof, that the SVT mechanism is AVNRT. The strength of that evidence can be increased if AOP is performed repeatedly and from different atrial sites, all yielding similar results. Variability in the postpacing VA interval after AOP from multiple distant atrial sites would be expected in the case of AT because the timing of the first return atrial impulse will depend on the proximity of the pacing site to the AT origin, and not on the timing of the first return ventricular beat. One small study suggested that first return VA intervals all within 14 ms of each other after “differential AOP” (AOP from two or three atrial sites: right atrial appendage, coronary sinus [CS] os, and distal CS) is consistent with a diagnosis of AVNRT (or AVRT) while VA interval differences obtained after differential AOP exceeding 14 ms is consistent with a diagnosis of septal AT.19 Typical AVNRT versus JT The only certain way to distinguish typical AVNRT from JT would be to record from the limbs of an AVNRT circuit within the AVN and demonstrate fusion in those recordings during resetting or entrainment of AVNRT, which would not occur in JT since its mechanism is not reentry. At present, this is not possible. Fortunately, typical AVNRT is both much more common than JT and is strongly favored when there is other evidence of dual AVN pathway physiology. In particular, AVNRT is strongly favored when the initiation of SVT appears to require a “jump” to the slow AVN pathway. Nevertheless, not all AVNRTs have demonstrable discontinuities in their AV nodal refractory curves and JT could appear to require a critical Atrio-His (AH) interval for its initiation by coincidence. An AH response would be expected after VOP in the case of either tachycardia. AOP is helpful in distinguishing AVNRT from JT.20 The last atrial paced beat would be expected to conduct with a long AH interval (slow AV nodal pathway conduction) to the last ventricular electrogram that is accelerated to the pacing CL before SVT resumes (Fig. 5). A prospective study has recently confirmed that this is the case.21 The obvious pitfall for this maneuver would be the exceptional circumstance where JT coexists with dual AVN physiology and where the last paced beat conducts to the ventricles via the slow AVN pathway and echoes back to the atria via the fast AVN pathway before JT resumes. As we will see, the coexistence of JT with dual AVN physiology is a common caveat for pacing maneuvers employed to distinguish AVNRT from JT. June 2012 763 VEENHUYZEN, ET AL. Figure 5. Atrial overdrive pacing to distinguish AVNRT from AT and JT. The termination of high right atrial pacing (from the electrode pair labeled ABLd) at a cycle length of 340 ms during SVT at a CL of 350 ms is shown, revealing the continuation of SVT after pacing stops. Atrial and ventricular activation during SVT is virtually simultaneous. The star indicates the last entrained His and ventricular electrograms and QRS complex. This beat and subsequent beats of the SVT demonstrate “VA linking”: the atrial activation sequence and VA interval on the last entrained beat is the same as during the tachycardia, suggesting that ventricular and atrial activation are mechanistically linked, which would not be expected if the diagnosis were atrial tachycardia (AT). During pacing, the PR interval exceeds the RR interval. This is consistent with antegrade conduction over a slow AV node pathway during pacing, which would not be the expected if the diagnosis were junctional tachycardia (JT). The AH interval, including the last entrained AH interval, is long, and the tachycardia resumes as the last entrained impulse echoes back up the fast AV node pathway. Again, this would not be expected if the diagnosis were JT. Padanilam and colleagues suggested that scanning diastole with atrial premature beats (APBs) can often be helpful to distinguish AVNRT from JT.22 A His-refractory APB (HRAPB) that affects the timing of the next His potential in any way (i.e. that advances or delays the next His potential, or that terminates the SVT) is consistent with a diagnosis of AVNRT (Fig. 6). An HRAPB should not be able to reach the AVN focus of a JT if retrograde conduction from that focus proceeds with roughly the same timing as antegrade conduction to the His bundle. The timing of His bundle depolarization is actually a surrogate for the timing of retrograde fast AVN pathway conduction. An APB that occurs when the fast pathway is refractory cannot affect a 764 JT focus since the stimulated wavefront would collide with the JT wavefront in or proximal to the fast AVN pathway. On the other hand, an APB that occurs when the FP is refractory can affect an AVNRT circuit by engaging the slow AVN pathway. Accordingly this response is specific for AVNRT. As with AOP, the coexistence of JT with dual AVN physiology represents a caveat since it could permit an HRAPB to advance (but not delay) the timing of the next His bundle depolarization if conducted via the slow pathway, leading to an echo beat via the fast pathway, only to have JT resume afterwards. Hamdan and colleagues have suggested that resetting by an APB delivered close to the AVN slow pathway region at a time when the June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 Figure 6. PAC response in AVNRT. Panel A: During supraventricular tachycardia with cycle length of 412 ms and simultaneous atrial and ventricular activation, a paced atrial premature beat (APB) is delivered by electrodes at the ostium of the coronary sinus (pCS) timed to junctional (His) refractoriness. The arrow points to the local atrial activation (A) occurring at the expected time of His bundle depolarization ([H]). Delay of the subsequent His by 16 ms indicates a diagnosis of atrioventricular nodal reentry tachycardia (AVNRT) where the APB prematurely activated the slow AVN pathway, delaying the subsequent beat due to decremental conduction slowing. Panel B: Schematic depiction of this response in AVNRT—the paced wavefront (square wave & solid arrow) can enter the excitable gap in the AV nodal circuit and activate the slow pathway. Delay of the subsequent His timing (as shown in Panel A) should be specific for AVNRT. Advancement of the subsequent His timing, or termination of the SVT, are also very specific for AVNRT, but could be observed in the case of JT with dual AV node physiology. Panel C: Schematic depiction of the response to an APB timed to junctional refractoriness in junctional tachycardia (JT); in the absence of dual AV nodal pathways the paced wavefront will collide with the retrograde wavefront from the JT focus (star) somewhere in the AV node or proximal conduction system (black bar), so His timing cannot be affected. septum is being depolarized indicates a diagnosis of AVNRT, but the same caveat discussed earlier would apply if dual AVN physiology coexisted with JT.23 Padanilam and colleagues also suggested that the continuation of SVT after advancement of His bundle depolarization by an APB delivered prior to His bundle depolarization (which is again acting as a surrogate for fast pathway depolarization) PACE, Vol. 35 identifies a diagnosis of JT.22 An APB delivered prior to fast pathway depolarization can advance the immediate His bundle depolarization via antegrade conduction down the fast pathway, but that would leave the fast pathway refractory and unable to participate in an ongoing AVNRT circuit (Fig. 7). Thus, both JT and AVNRT could terminate when an APB that occurs prior to His bundle depolarization advances the immediate His potential, but only JT would be expected June 2012 765 VEENHUYZEN, ET AL. Figure 7. PAC response in JT. Panel A: During supraventricular tachycardia (SVT) with a cycle length of 560 ms and simultaneous atrial and ventricular activation, a paced atrial premature beat (APB) is delivered by electrodes at the high right atrium (HRA) prior to anticipated junctional (His) refractoriness (arrow), advancing the immediate His by 38 ms, and the SVT continues, indicating a diagnosis of junctional tachycardia (JT). Panel B: Schematic depiction of this response in JT; an early paced APB can advance the timing of His activation, while resetting the JT focus (faded star), after which the JT will resume. Panel C: Schematic depiction of the response to an early paced APB in AVNRT; the paced wavefront may advance the subsequent His timing by conducting via the AV nodal fast pathway, but will also collide with the AVNRT circuit in the AV nodal slow pathway (black bar) or leave the fast pathway refractory, terminating the tachycardia. to be able to continue in these circumstances (Fig. 7). The caveat to this situation again includes the coexistence of dual AVN physiology with JT where the early APB results in a so-called “dual response.” The immediate His potential is advanced via conduction down the fast pathway, but simultaneous conduction down the slow pathway resets an AVNRT circuit. 766 AV-dissociated SVT The differential diagnosis of an AVdissociated SVT includes AVNRT with block to the atrium, JT with block to the atrium, orthodromic nodoventricular reciprocating tachycardia (ONVRT), and orthodromic nodofascicular reciprocating tachycardia (ONFRT). Intra-Hisian reentry has also been June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 proposed, but we are not aware of any cases demonstrating the existence of this mechanism, and it will not be discussed further. (Of course, ventricular tachycardia, including bundle branch reentry, would have to be excluded, but the means to do so are beyond the scope of this review.) The ability of an HRVPB to terminate the SVT or to advance/delay (reset) the timing of the subsequent His potential would exclude AVNRT and JT. Resetting by a VPB and entrainment by VOP with evidence of fusion would exclude AVNRT, JT, and ONFRT, because fusion would specifically indicate collision of the stimulated antidromic wavefront with the orthodromic wavefront from the preceding beat occurring in ventricular myocardium (Fig. 8).24 Figure 8. His-refractory ventricular premature beats (HRVPBs) in AV-dissociated SVT. HRVPBs are delivered from the right ventricular apex (RVA) during a tachycardia with a typical right bundle branch block (RBBB) QRS morphology and a prolonged His-ventricular interval (Panels A and B). Independent atrial activity (A) is seen, indicating atrioventricular dissociation. The main differential diagnosis for this AV-dissociated tachycardia includes (1) AVNRT, (2) junctional tachycardia (JT), (3) intra-Hisian reentry, (4) bundle branch reentry VT (BBRVT), and (5) a orthodromic reentry using a nodoventricular or nodofascicular connection as the retrograde limb (with retrograde block to the atrium in each case). In Panels (A) and (B), the HRVPBs advance the next H and V, ruling out possibilities (1), (2), and (3) since the paced retrograde wavefront would be unable to reach the circuits or ectopic focus if the His bundle had just been depolarized. These circumstances are depicted in Panels C (for AVNRT) and D (for JT, but would equally apply for intra-Hisian reentry). Panels E and F show the circuit for BBRVT and reentry using a nodofascicular connection, respectively. An earlier PVC (Panel B) leads to a 20-ms increase in the stimulus-to-His interval compared to Panel A, suggesting decremental conduction in the circuit somewhere between the ventricle and the His. This would be expected if the AV node was part of the circuit (Panel F) but would be unusual if it was BBRVT (Panel E). Note that the HRVPB in Panel (A) is fused (compare to the QRS complex morphology of the HRVPB in Panel B), indicating that collision of the stimulated orthodromic wavefront with the antidromic wavefront from the preceding beat has occurred in ventricular myocardium. This is not possible during orthodromic nodofascicular reentry, where this collision would be expected to occur within the conduction system. The best explanation for these findings is orthodromic nodoventricular reciprocation (see Ref. 23). PACE, Vol. 35 June 2012 767 VEENHUYZEN, ET AL. As with conventional orthodromic AVRT circuits, basal pacing sites close to the ventricular insertion of the nodoventricular AP would be expected to increase the likelihood of detecting fusion, and also decrease the postpacing interval (PPI)-TCL difference and could, theoretically, be used as a method of mapping the ventricular insertion of the AP; sites closest to the ventricular insertion would be expected to have the shortest PPI-TCL difference (provided that decremental conduction through the AVN does not influence the result, so the VOP CL should be as consistent as possible) and a QRS complex morphology closest to that of the native SVT, possibly even revealing entrainment with concealed fusion. Haı̈ssaguerre and colleagues have described the use of single ventricular extrasystoles to accomplish the same discriminatory goals.25 It is noteworthy that ONVRT and ONFRT need not be AV dissociated. Indeed, if AV associated, either could have any VA interval, including one short enough to mimic typical AVNRT. Clues to their presence may include an SVT that otherwise appears consistent with AVNRT but where: (1) VOP yields cPPI-TCL and SA-VA interval values that are too short to be consistent with AVNRT; (2) VOP results in manifest entrainment (which should not be possible in AVNRT or ONFRT, therefore indicating a diagnosis of ONVRT); or (3) HRVPBs terminate the SVT or affect the timing of the next His potential. The latter is the only reason we can think of to bother scanning diastole with VPBs during an SVT that appears consistent with typical AVNRT. AVNRT with Eccentric Left Atrial Activation AVNRT with eccentric left atrial activation is an uncommon SVT whose existence is well documented.26–32 Although it is widely recognized that the atrial exit of the AVN is not always in the superior septum but may also be in the inferior septum where it can extend along the left inferior aspect of the mitral annulus, it is not as well recognized that AVNRT can rarely have its earliest atrial activation along the lateral mitral annulus. Normally, this atrial activation sequence would lead one to exclude AVNRT and to consider AVRT employing a left free wall AP or a left-sided AT. If one entertains the possibility of AVNRT with a left atrionodal extension in the differential diagnosis, then one faces a similar diagnostic challenge to that posed by SVT with concentric atrial activation: all SVT mechanisms are possible. The usual ventricular pacing maneuvers may be helpful, but the yield is likely to be lower if they are only performed from the RV apex, which is relatively far from the lateral mitral annulus. As 768 discussed earlier, an HRVPB may not be capable of resetting AVRT employing a left free wall AP because the stimulated wavefront may not have enough time to reach the operative AP if delivered late enough to be His refractory. In the setting of AVRT employing a left free wall AP, VOP is unlikely to result in manifest entrainment, and the cPPI-TCL and SA-VA interval differences may be long by virtue of the distance of the RV apical pacing site from the circuit (this would not be the case if there was left bundle branch block during SVT, as the RV would be part of the circuit in this circumstance). Thus, if the usual ventricular pacing maneuvers are not helpful, one should consider scanning diastole with VPBs delivered from basal sites in the left ventricle (LV) or performing VOP from basal sites in the LV (which can be stimulated via a branch of the CS, obviating the need to access the systemic circulation prior to making a diagnosis in at least one-third of cases). Alternatively, “differential entrainment,” as described in Part 1 of this review, could be performed.1 It is noteworthy that differential entrainment has only been studied in a few cases of AVRT employing left free wall APs and we consider it at least theoretically possible that a patient could have a considerably shorter conduction time from the RV apex to a left posterior AP than from the basal infundibulum to such an AP, potentially making the results of differential entrainment misleading. Conclusion In Part 1 of this review, we explored how attempts to continuously reset (i.e. entrain) SVT by VOP can be used to provide a diagnosis in the majority of sustained, regular SVTs. In this part, we have explored other diagnostic pacing maneuvers that might be helpful when VOP is not diagnostic or appropriate including attempts to reset SVT with single atrial or ventricular beats, para-Hisian pacing, apex versus base pacing, and AOP. We have also discussed some specific diagnostic SVT challenges encountered in the electrophysiology lab. To be sure, there are other pacing maneuvers that we have not addressed, but this review should serve as a thorough foundation for SVT diagnosis and for understanding the strengths and weaknesses of those other maneuvers. We hope that by gaining a thorough understanding of how these maneuvers exploit differences in the underlying anatomy and physiology of the various SVT mechanisms, one will gain an appreciation of which diagnostic pacing maneuvers constitute “proof” and which are merely “evidence” in favor of one mechanism or another. June 2012 PACE, Vol. 35 DIAGNOSTIC SVT PACING MANEUVERS 2 References 1. Veenhuyzen GD, Quinn FR, Wilton SB, Clegg R, Mitchell LB. Diagnostic pacing maneuvers for supraventricular tachycardia: Part 1. Pacing Clin Electrophysiol 2011; 34:767–782. 2. Josephson ME. Supraventricular tachycardias. In: Josephson ME (ed): Clinical Cardiac Electrophysiology. Philadelphia, PA, Lea and Febiger; 1993, p. 242. 3. Boonyapisit W, Kuhne M, Morady F, Jongnarangsin K, Supraventricular tachycardia with delayed atrial depolarization following Hissynchronous ventricular stimulation: What is the mechanism? Heart Rhythm 2010; 7:280–281. 4. Matsushita T, Badhwar N, Collins KK, Van Hare GF, Barbato G, Lee BK, Lee RJ, et al. Usefulness of a ventricular extrastimulus from the summit of the ventricular septum in diagnosis of septal accessory pathway in patients with supraventricular tachycardia. Am J Cardiol 2004; 93:643–646. 5. Anselme F. Minimal/essential electrophysiologic assessment before ablation. In: Zipes DP, Haissaguerre M. (eds.): Catheter Ablation of Arrhythmias, 2nd Ed. Armonk, New York, Futura Publishing Company, Inc; 2002, p. 38. 6. Hsu JC, Tanel RE, Lee BK, Scheinman MM, Badhwar N, Lee RJ, Tseng ZH, et al. Differences in accessory pathway location by sex and race. Heart Rhythm 2010; 7:52–56. 7. Hirao H, Otomo K, Wang X, Beckman KJ, McClelland JH, Widman L, Gonzalez MD, et al. Para-Hisian pacing. A new method for differentiating retrograde conduction over an accessory AV pathway from conduction over the AV node. Circulation 1996; 94:1027–1035. 8. Matsushita T, Hongo RH, Badhwar N, Scheinman MM. Define the mechanism of the tachycardia and explain the results of para-Hisian pacing. J Cardiovasc Electrophysiol 2004; 15:504–505. 9. Takatsuki S, Mitamura H, Tanimoto K, Fukuda Y, Ieda M, Miyoshi S, Soekima K, et al. Clinical implications of “pure” Hisian pacing in addition to para-Hisian pacing for the diagnosis of supraventricular tachycardia. Heart Rhythm 2006; 3:1412–1418. 10. Heidbuchel H, Ector H, Adams J, Van de Werf F. Use of only a regular diagnostic His-Bundle catheter for both fast and reproducible “paraHisian pacing” and stable right ventricular pacing. J Cardiovasc Electrophysiol 1997; 8:1121–1132. 11. Adachi M, Igawa O, Miyake J, Yano A, Inoue Y, Ogura K, Kato M, et al. QRS complex widening due to loss of left bundle branch capture: Pitfall of para-Hisian pacing. J Interv Card Electrophysiol 2009; 25:213–216. 12. Obeyesekere M, Leong-Sit P, Skanes A, Krahn A, Yee R, Gula LJ, Bennett M, et al. Determination of inadvertent atrial capture during para-Hisian pacing. Circ Arrhythm Electrophysiol 2011; 4:510–514. 13. Reddy VY, Jongnarangsin K, Albert CM, Sabbour H, Keane D, Mela T, McGovern B, et al. Para-Hisian entrainment: A novel pacing maneuver to differentiate orthodromic atrioventricular reentrant tachycardia from atrioventricular nodal reentrant tachycardia. J Cardiovasc Electrophysiol 2003; 12:1321–1328. 14. Martinez-Alday JD, Almendral J, Arsenal A, Ormaetxe JM, Pastor A, Villacastı́n JP, Medina O, et al. Identification of concealed posteroseptal Kent pathways by comparison of ventriculoatrial intervals from apical and posterobasal right ventricular sites. Circulation 1994; 89:1060–1067. 15. Segal OR, Gula LJ, Skanes AC, Krahn AD, Yee R, Klein GJ. Differential ventricular entrainment: A maneuver to differentiate atrioventricular node reentrant tachycardia from orthodromic reciprocating tachycardia. Heart Rhythm 2008; 6:493–500. 16. Knight BP, Ebinger M, Oral H, Kim MH, Sticherling C, Pelosi F, Michaud GF, et al. Diagnostic value of tachycardia features and pacing maneuvers during paroxysmal supraventricular tachycardia. J Am Coll Cardiol 2000; 36:574–782. 17. Yamada T, Huizar JF, McElderry HT, Kay GN. Atrial tachycardia with slow pathway conduction mimicking typical atrioventricular nodal reentrant tachycardia. Europace 2007; 9:299–301. PACE, Vol. 35 18. Gula LJ, Skanes A, Krahn AD, Klein GJ. Novel approach to diagnosis of a wide-complex tachycardia. J Cardiovasc Electrophysiol 2004; 15:466–469. 19. Maruyama M, Kobayashi Y, Miyauchi Y, Ino T, Atarashi H, Katoh T, Mizuno K. The VA relationship after differential atrial overdrive pacing: A novel tool for the diagnosis of atrial tachycardia in the electrophysiologic laboratory. J Cardiovasc Electrophysiol 2007; 18:1127–1133. 20. McElderry, Kay GN. Ablation of atrioventricular nodal reentry by the anatomic approach. In: Huang S, Wood MA (eds.): Catheter Ablation of Cardiac Arrhythmias. Philadelphia, OA, Saunders Elsevier, 2006, p. 334. 21. Fan R, Tardos JG, Almasry I, Barbera S, Rashba EJ, Iwai W. Novel use of atrial overdrive pacing to rapidly differentiate junctional tachycardia from atrioventricular nodal reentrant tachycardia. Heart Rhythm 2011; 8:840–844. 22. Padanilam BJ, Manfredi JA, Steinberg LA, Olson JA, Fogel RI, Prystowsky EN. Differentiating junctional tachycardia and atrioventricular node re-entry tachycardia based on response to atrial extrastimulus pacing. J Am Coll Cardiol 2008; 52:1711–1717. 23. Hamdan MH, Page RL, Scheinman MM. Diagnostic approach to narrow complex tachycardia with VA block. Pacing Elecrophysiol 1997; 20:2984–2988. 24. Quinn FR, Mitchell LB, Mardell AP, Disler D, Veenhuyzen GD. Entrainment mapping of a concealed nodoventricular accessory pathway in a man with complete heart block and tachycardia induced cardiomyopathy. J Cardiovasc Electrophysiol 2008; 19:90–94. 25. Haı̈ssaguerre M, Campos J, Marcus FI, Papouin G, Clementy J. Involvement of a nodofascicular connection in supraventricular tachycardia with VA dissociation. J Cardiovasc Electrophysiol 1994; 5:854–862. 26. Dizon J, Reiffel J, Kassotis J, Woollett, Garan H. Change in the retrograde atrial activation sequence following radiofrequency modification of the atrioventricular node: Implications for the electrophysiologic circuit of a variant of atrioventricular node reentrant tachycardia. J Cardiovasc Electrophysiol 2003; 14:461–466. 27. Hwang C, Martin DJ, Goodman JS, Gang ES, Mandel WJ, Swerdlow CD, Peter CT, et al. Atypical atrioventricular node reciprocating tachycardia masquerading as tachycardia using a left sided accessory pathway. J Am Coll Cardiol 1997; 30: 218–225. 28. Nam GB, Rhee KS, Kim J, Choi KJ, Kim YH. Left atrionodal connections in typical and atypical atrioventricular nodal reentrant tachycardias: Activation sequence in coronary sinus and results of radiofrequency catheter ablation. J Cardiovasc Electrophysiol 2006; 17:1–7. 29. Otomo K, Okamura H, Noda T, Satomi K, Shimizu W, Kurita T, Aihara N, et al. “Left variant” atypical atrioventricular nodal reentrant tachycardia: Electrophysiological characteristics and effect of slow pathway ablation within coronary sinus. J Cardiovasc Electrophysiol 2006; 17:1184–1186. 30. Anselme F, Papageorgiou P, Monahan K, Zardini M, Boyle N, Epstein LM, Josephson ME. Presence and significance of the left atrionodal connection during atrioventricular nodal reentrant tachycardia. Am J Cardiol 1999; 83:1530–1536. 31. Chen J, Anselme F, Smith TW, Zimetbaum P, Epstein LM, Papageorgiou P, Josephson ME. Standard right atrial ablation is effective for atrioventricular nodal reentry with earliest activation in the coronary sinus. J Cardiovasc Electrophysiol 2004; 15:2–7. 32. Ong MG, Lee PC, Tai CT, Lin YJ, Hsieh MH, Chen YJ, Lee KT, et al. The electrophysiologic characteristics of atrioventricular nodal reentry tachycardia with eccentric retrograde activation. Int J Cardiol 2007; 120:115–122. June 2012 769