Paroxysmal supraventricular tachycardia with Wenckebach block: evidence for reentry within the upper portion of the atrioventricular node

Comparison of adenosine effects on atrioventricular node reentry and atrioventricular reciprocating tachycardias

BACKGROUND

Adenosine is an established first line therapy for the treatment of narrow complex tachycardias. The two most common etiologies of paroxysmal supraventricular tachycardia (SVT) are atrioventricular node reentry tachycardia (AVNRT) and atrioventricular reciprocating tachycardia (AVRT).

HYPOTHESIS

We postulated that adenosine might have different effects on the termination of AVNRT vs. AVRT, and that these differences might assist in the noninvasive differentiation between these diagnoses.

METHODS

Fifty-nine patients referred for the diagnosis and treatment of SVT were included in the study. All patients had SVT induced during electrophysiology testing, and each patient received adenosine during SVT. The adenosine dose, time to tachycardia termination, and site of tachycardia termination were recorded. Seventeen patients required isoproterenol administration to initiate SVT. This subset of patients was compared with those not requiring isoproterenol.

RESULTS

There was no statistically significant difference in the adenosine dose or time to tachycardia termination when comparing patients with AVNRT with those with AVRT. All patients with AVNRT had termination of tachycardia in the antegrade direction with final activation in the atria. Patients requiring isoproterenol for tachycardia initiation experienced tachycardia termination significantly faster than those not requiring isoproterenol, although there was no difference in the dose of adenosine required for termination.

CONCLUSION

These data demonstrate that patients with dual AV node physiology and AVNRT do not have altered sensitivity to adenosine compared with patients with AVRT and normal AV nodes. Further investigation will be required to determine the clinical utility of the significantly shorter time to tachycardia termination for patients receiving isoproterenol.

Unusual Wenckebach phenomenon due to an atrial tachycardia arising from the apex of Koch's triangle in the presence of dual AV nodal physiology

A case of a patient with narrow QRS tachycardia and without structural heart disease is presented. The electrophysiologic study revealed an atrial tachycardia in the presence of dual atrioventricular (AV) nodal physiology and AV block at suprahisian level, the latter two leading to an unusual Wenckebach periodicity. The entire septal area was mapped as was the coronary sinus (CS) os and the earliest atrial activation was found at the apex of Koch's triangle in close vicinity to the His bundle (HB). Cryomapping at that point reproducibly terminated the tachycardia without impairing AV conduction. Cryoablation rendered the tachycardia non-inducible. Discontinuous AV conduction persisted but AV nodal reentrant tachycardia (AVNRT) was not inducible. Six months later the patient is arrhythmia-free.

Atrioventricular nodal reentrant tachycardia with advanced infra-hisian atrioventricular block

We report a case of 78-year-old man admitted to the hospital due to palpitations and lightheadedness. On EKG advanced atrioventricular block with ventricular rate of 37 beats per minute was noted. On electrophysiology study a common type of atrioventricular nodal reentrant tachycardia was inducible with maintenance of advanced AV block. Radiofrequency ablation of slow pathway followed by placement of a permanent pacemaker resulted in elimination of tachycardia and resolution of symptoms.

Location of the lower turnaround point in typical AV nodal reentrant tachycardia: a quantitative model

INTRODUCTION

Recent observations suggest that the circuit of AV nodal reentrant tachycardia (AVNRT) may extend down to the His bundle. The purpose of this study was to develop a quantitative model indicating the location of the lower turnaround point in AVNRT.

METHODS AND RESULTS

Slow pathway modification was performed in 70 patients with typical AVNRT. During sinus rhythm, ventricular pacing was performed with the AVNRT cycle length. During AVNRT, the HinitAinit interval was measured from initial His to the initial atrial deflection recorded in the His-bundle lead. During ventricular pacing, the HendAinit interval was measured from end of the His to the beginning of the atrial deflection. It was hypothesized that x reflects conduction time from the lower turnaround point to Ainit, whereas y reflects conduction time from the lower turnaround point to Hinit. Anterograde conduction during AVNRT and retrograde conduction during ventricular pacing were assumed to be identical if there was 1:1 retrograde conduction at the AVNRT cycle length. The following formulas describe the relation of the measured parameters: x - y = HinitAinit; and x + y = HendAinit. Resolving both formulas yields the unknown x and y: y = (HendAinit - HinitAinit)/2, x = (HendAinit + HinitAinit)/2. These criteria were present in 52 of 70 patients. The mean cycle length of AVNRT was 355 +/- 42 msec, mean HinitAinit was 54 +/- 27 msec, and mean HendAinit was 60 +/- 29 msec. Accordingly, in 20 of 52 patients, the lower turnaround point was located within the His bundle (y = -15.4 +/- 16.1 msec), in 3 of 52 it was in the nodal-His junctional area (y = 0), and in 29 of 52 it was above the His bundle (y = +12.7 +/- 10.3 msec). The HinitAinit interval was significantly longer (66 +/- 32 msec vs 47 +/- 20 msec; P = 0.02) and the HendAinit interval was significantly shorter (45 +/- 30 msec vs 69 +/- 24 msec; P = 0.004) when the first group was compared with the others.

CONCLUSION

In about 1 of 3 of patients with typical AVNRT, the lower turnaround point of the circuit is within the His bundle; in more than half of the patients it is above the His bundle. These data do not support the concept that all AVNRTs have an intranodal circuit, but are in accordance with the finding of longitudinal dissociation of the His bundle.

Entrainment of typical AV nodal reentrant tachycardia using para-Hisian pacing: evidence for a lower common pathway within the AV node

INTRODUCTION

Despite the ability to cure atrioventricular nodal reentrant tachycardia (AVNRT) by radiofrequency catheter ablation with a high success rate, the exact localization of the tachycardia circuit is still not well established. The presence of AV nodal tissue between the typical AVNRT circuit and the His bundle, constituting a lower common pathway (LCP), remains controversial.

METHODS AND RESULTS

Entrainment of AVNRT during para-Hisian stimulation allows accurate measurement of the His- to- atrial (HA) interval which is part of the same circuit as that of the tachycardia. With an LCP, during tachycardia, there is simultaneous conduction from the low turnaround of the circuit to the atrium (via the fast pathway) and to the His bundle (via the LCP). However, during entrainment by para-Hisian pacing, the impulse has to retrogradely depolarize sequentially the LCP and the fast pathway. Therefore, in the presence of an LCP, the HA interval duration during tachycardia (HAt) should be shorter than that of during entrainment by para-Hisian stimulation (HAe). We considered an LCP present when Hae - HAt was > or = 10 msec. Entrainment of typical AVNRT with para-Hisian stimulation was performed in 23 consecutive patients (21 females) with a mean age of 45+/-17 years. LCP was considered to be present in 18 of 23 patients (78%). In addition, transient His-bundle dissociation from the ongoing tachycardia occurred in seven patients (30%).

CONCLUSION

These results support the presence of a LCP during typical AVNRT.

Ventriculoatrial block during atrioventricular nodal reentrant tachycardia utilizing multiple retrograde pathways

A rare case of narrow QRS tachycardia continuing despite the occurrence of VA block is reported. Right ventricular stimulation suggested dual AV nodal physiology. The tachycardia was induced by ventricular premature stimulation, which failed to depolarize the atrium. Two types of tachycardia that had different retrograde conduction sequences, HA intervals, and cycle lengths were induced. The occurrence of VA block did not terminate the tachycardia but transiently prolonged the tachycardia cycle length. These findings suggest the mechanism is AV nodal reentry utilizing multiple retrograde pathways with intranodal reentry bridging the VA block and maintaining the tachycardia.

Atrioventricular nodal reentrant tachycardia with atrioventricular block

Atrioventricular block (AVB) during atrioventricular nodal reentrant tachycardia (AVNRT) has been well documented, although it is not a common phenomenon. The mechanism for the initiation and resolution of AVB during AVNRT have been postulated. However, the site of AVB and its implication on the reentrant circuit in AVNRT is not clear. We illustrate two examples of AVNRT with AVB and offer further clarification on the site and mechanism of AVB.

2:1 atrioventricular block during atrioventricular node reentrant tachycardia

OBJECTIVES

The purpose of this study was to determine the incidence and to clarify the mechanism of 2:1 atrioventricular (AV) block during AV node reentrant tachycardia induced in the electrophysiology laboratory.

BACKGROUND

In patients with 2:1 AV block during AV node reentrant tachycardia, the absence of a His bundle potential in the blocked beats has been considered evidence of intranodal, lower common pathway block.

METHODS

In consecutive patients with AV node reentrant tachycardia, the incidence of 2:1 AV block and the response to atropine and a single ventricular extrastimulus was observed.

RESULTS

Persistent 2:1 AV block occurred in 13 of 139 patients with AV node reentrant tachycardia. A His bundle deflection was present in the blocked beats in eight patients and absent in five. Patients with 2:1 AV block had a shorter tachycardia cycle length than did patients without such block (mean +/- SD 312 +/- 32 vs. 353 +/- 55 ms, p < 0.01). Atropine did not alter the 2:1 block in any patient. In every patient, a single ventricular extrastimulus introduced during the tachycardia converted the 2:1 block to 1:1 conduction.

CONCLUSIONS

The incidence of induced 2:1 AV block during AV node reentrant tachycardia is approximately 10%. The lack of a response to atropine and the consistent conversion of 2:1 block to 1:1 conduction by a ventricular extrastimulus indicate that, regardless of the presence or absence of a His bundle potential in blocked beats, 2:1 block during AV node reentrant tachycardia is due to functional infranodal block.

Elimination of AV nodal reentrant tachycardia with 2:1 VA block by posteroseptal ablation

AV nodal reentry capable of VA block during tachycardia was successfully eliminated using a posteroseptal ablation pulse delivered well away from the site of earliest atrial activation during tachycardia. A possible explanation is that the arrhythmia represented typical AV nodal reentrant tachycardia with transient intra-atrial conduction block during tachycardia.

"AV nodal" reentry: Part II: AV nodal, AV junctional, or atrionodal reentry?

The classical model of "atrioventricular (AV) nodal" reentrant tachycardia suggests that the reentrant circuit is entirely within the compact AV node and that AV nodal tissue is present proximal and distal to the circuit. Recent evidence from mapping studies and from examination of the effects of curative procedures, however, suggests that the upper end of the circuit uses perinodal atrial or transitional tissue. Moreover, the anatomical substrate of dual "AV nodal" pathways is likely to be the multiple connections between compact AV node and atrium rather than discrete intranodal pathways. The antegrade slow pathway appears to be situated at the posteroinferior approaches to the AV node in the region between the coronary sinus orifice and the tricuspid annulus. The retrograde fast pathway appears to be situated in the anterior atrionodal connections at the apex of Koch's triangle, close to the His bundle. The lower turnaround point of the circuit is likely to be within the AV node.

Complexities of junctional tachycardias

The atrioventricular junction is a compact area in which most of the known electrophysiologic substrates and mechanisms play a role in the genesis and maintenance of tachyarrhythmias. The purpose of this review is to summarize the data on normal atrioventricular junction anatomy and electrophysiologic function and correlate that information with surface electrocardiographic recordings, intracardiac electrophysiologic data, and interventional data from surgical and catheter techniques. Models of tachycardia mechanisms are proposed for typical and atypical atrioventricular nodal reentrant tachycardia, permanent junctional reciprocating tachycardia, and orthodromic supraventricular tachycardias utilizing "intermediate septal" accessory connections.

Radiofrequency ablation for atrioventricular node reentrant tachycardia: comparison between fast (anterior) and slow (posterior) pathway ablation

OBJECTIVES

We compared the electrophysiologic effects on atrioventricular (AV) node physiology of selective "fast" versus selective "slow" pathway radiofrequency ablation in 42 patients with drug-resistant AV node reentrant tachycardia who underwent 51 ablation attempts to prevent tachycardia recurrence while preserving AV conduction.

BACKGROUND

The recent introduction of radiofrequency ablation to treat AV node reentrant tachycardia allows the opportunity to study the effects of selective elimination of the different limbs involved in AV node reentrant tachycardia.

METHODS

Selective fast pathway ablation was attempted in 13 patients by delivering radiofrequency energy anteriorly across the tricuspid valve anulus. Selective slow pathway ablation was attempted in 29 patients by delivering radiofrequency energy posteriorly across the tricuspid valve anulus at sites where putative slow pathway potentials were recorded.

RESULTS

Selective fast pathway ablation eliminated AV node reentrant tachycardia without AV block in 6 (46%) of 13 patients after one ablation session and in an additional 3 patients (69% of total) after repeat ablation sessions. Slow pathway ablation eliminated AV node reentrant tachycardia without AV block in 26 (90%) of 29 patients after one radiofrequency ablation session and in an additional 2 patients (97% of total) after repeat ablation sessions. Selective fast pathway ablation increased the PR interval (140 to 220 ms, p = 0.0001) and AH interval (66 to 153 ms, p = 0.0001), whereas slow pathway ablation did not change these intervals. Fast pathway radiofrequency ablation caused retrograde block in 7 (64%) of 11 patients, whereas no patients undergoing slow pathway ablation developed selective retrograde block. Single AV node echo beats were commonly induced after slow but not fast pathway ablation (17 of 29 patients vs. 1 of 11 patients, respectively, p = 0.01) and did not predict recurrence of AV node reentrant tachycardia.

CONCLUSIONS

Successful selective radiofrequency ablation of fast or slow pathways in patients with AV node reentrant tachycardia resulted in different electrophysiologic properties after ablation. Slow pathway ablation produced more successful outcomes, with a decreased prevalence of recurrent AV node reentrant tachycardia or AV block.

Recovery of retrograde fast pathway excitability in the atrioventricular node reentrant circuit after concealed anterograde impulse penetration

The recovery of the retrograde fast pathway excitability in atrioventricular (AV) node reentry has been difficult to assess with ventricular extrastimulation because of difficulty in achieving sufficiently short intranodal coupling intervals and the potential interposition of "lower common pathway" nodal tissue. To circumvent these methodologic obstacles in 10 patients with inducible AV node reentrant tachycardia, a fixed atrial extrastimulus (A2) coupled to a basic atrial drive (A1) at a cycle length of 500 ms was utilized to reproducibly initiate AV node reentrant echoes. A ventricular extrastimulus (V3) was then introduced after A2 at progressively shorter coupling intervals (A2V3) in an attempt to pre-excite the retrograde fast pathway after concealed anterograde penetration by A2. In six patients, retrograde fast pathway pre-excitation was achieved at critical A2V3 intervals, as evidenced by the appearance of A3 by up to 28 +/- 6 ms in advance of the expected first AV node reentrant echo. In five of the six cases, the V3A3 interval was virtually unaltered (less than or equal to 5 ms decrease) when A2 was omitted. In seven patients, at a critically short A2V3 coupling interval (195 +/- 27 ms ), V3 abruptly failed to elicit A3 and concomitantly abolished all AV node echoes; yet when A2 was omitted, an A3 response returned, with V3A3 identical to previous values.(ABSTRACT TRUNCATED AT 250 WORDS)

Discordant effects of carotid sinus massage and intravenous adenosine in atypical (fast-slow) atrioventricular nodal reentrant tachycardia

The precise mechanism underlying supraventricular tachycardia with a long R-P interval is often difficult to assess noninvasively. Carotid sinus massage has been used traditionally to produce transient AV nodal conduction delay at the bedside, and may be of diagnostic or therapeutic benefit. More recently, adenosine has also been shown to be useful in this situation. We report the case of a patient with an incessant long R-P tachycardia in whom the response to CSM was misleading while the response to adenosine was diagnostic. The electrophysiologic responses to both maneuvers are displayed, and a mechanism for the discordant responses is proposed.

Double atrial and double ventricular responses during slow-fast fast-slow atrioventricular nodal reentrant tachycardia

A case was described with fast-slow form of atrioventricular nodal reentrant tachycardia as related with simultaneous fast and slow pathway conduction both antegrade and retrograde. Fast-slow form of tachycardia was induced by premature right atrial stimulation or incremental right ventricular pacing when the last paced beat conducted to the atria via both fast and slow pathways of the atrioventricular node causing double atrial response. Fast-slow form of tachycardia was spontaneously shifted to slow-fast form when the atrial echo, possibly through the retrograde intermediate pathway, was conducted antegradely over the fast and slow pathways simultaneously, producing double ventricular response.

The excitable gap in atrioventricular nodal reentrant tachycardia. Characterization with ventricular extrastimuli and pharmacologic intervention

Our purpose was to characterize the excitable gap during atrioventricular nodal reentrant tachycardia (AVNRT) to elucidate the electrophysiologic substrate of this clinically familiar microreentrant arrhythmia. Accordingly, in 11 patients with classic slow-fast AVNRT (mean cycle length, 342 +/- 41 msec), a single ventricular extrastimulus (V2) was periodically delivered after a spontaneous tachycardia beat (V1) until ventricular refractoriness was reached. With this technique, an excitable gap was considered present when atrial preexcitation of at least 20 msec could be achieved along with tachycardia resetting (noncompensatory pause after V2). The range of V1V2 intervals that resulted in atrial preexcitation constituted the preexcitation zone. Five patients (45%) showed evidence of an excitable gap at baseline, with a maximal atrial preexcitation achievable of 33 +/- 6 msec, representing 9 +/- 1% of the tachycardia cycle length. Verapamil was then administered to all 11 patients with the purpose of slowing the anterograde tachycardia wavefront before arrival of V2. This resulted in widening of the preexcitation zone in three patients by a mean of 50 +/- 37 msec, with a corresponding increase in maximal atrial preexcitation to 70 +/- 32 msec, or 16 +/- 4% of AVNRT cycle length, and the appearance of atrial preexcitation in two patients who lacked it during baseline. In the remaining six patients, AVNRT was not sustained after verapamil or was too unstable for evaluation. During baseline, V2A2 conduction time increased by only 5 +/- 3 msec throughout the preexcitation zone, with values at the outer border unchanged after verapamil, implying a fully excitable gap in the retrograde limb. In all patients with a preexcitation zone, AVNRT was consistently reset by V2, both at baseline and after verapamil, with a "flat" but mainly "increasing" response pattern as V1V2 was shortened. Hence, a significant number of patients with AVNRT have evidence of an excitable gap whose demonstrability can be facilitated by pharmacologic intervention; documentation of an increasing resetting response pattern, most apparent after verapamil, provides new evidence for a reentrant mechanism in AVNRT; and while not definitively proven, the presence of a fully excitable gap during AVNRT is most consistent with a microreentry circuit that incorporates an anatomic obstacle.

Fast-slow type of atrioventricular nodal reentrant tachycardia: horizontal dissociation of the AV node during tachycardia

Two patients with recurrent supraventricular tachycardia are presented. The tachycardia was initiated and terminated by atrial extrastimulation beyond the atrial relative refractory period and the atrial activation sequence during the tachycardia was low to high. The induction of tachycardia was dependent on a critical AH interval. In patient 1 who had ventriculoatrial conduction, the tachycardia was initiated by the premature ventricular stimulation followed by double atrial response. In patient 2 the ventriculoatrial conduction was not observed. In both patients, the unchanged atrial cycle length during the tachycardia with antegrade Wenckebach AH block was observed. When AH block occurred during tachycardia the first AH interval was shorter than the subsequent HA interval. In patient 2 verapamil (5 mg) prolonged the atrial cycle length during tachycardia and rapid intravenous injection of adenosine triphosphate (10 mg) terminated the tachycardia. Oral diltiazem (180 mg/day) suppressed the tachycardia in patient 1. These findings suggest that the mechanism of tachycardia may be fast-slow type of AV nodal reentry in the upper portion of the AV node and this type of arrhythmia has tendency to show incessant form.

Cryosurgical ablation of atrioventricular nodal reentry: histologic localization of the proximal common pathway

A method using cryosurgery has been previously described to selectively ablate atrioventricular nodal reentry tachycardia while preserving intact atrioventricular conduction. The purpose of the present study was to define the histologic features of the cryolesions in relationship to the specialized conduction system. In 12 adult dogs a series of nine discrete cryolesions was placed along the perimeter of the triangle of Koch while continuously monitoring the His bundle electrogram. All animals survived the operation and maintained intact atrioventricular conduction. At 14 weeks after surgery the hearts were sectioned and examined. In all 12 animals there was a confluent mass of dense fibrous tissue present in the lower atrial septum that was in immediate proximity to but did not involve the atrioventricular node-His bundle. The ablation of perinodal tissue with preservation of the specialized conduction system with the use of this cryosurgical technique was confirmed. It is likely that the cryoablated perinodal tissue represents the proximal common pathway of the circuit for atrioventricular nodal reentry tachycardia.

Cryosurgical treatment of atrioventricular node reentrant tachycardia

Paroxysmal supraventricular tachycardia most commonly arises from reentry within the atrioventricular (AV) node. Although ablation of the His bundle has gained popularity for treating patients with AV node reentrant tachycardia refractory to medical therapy, undesirable sequelae include complete heart block and the necessity for a permanent pacemaker. To obviate this limitation, we have developed a discrete cryosurgical procedure that interrupts the reentrant circuit responsible for AV node reentrant tachycardia without blocking AV conduction. After first characterizing the salutary effects of this approach in experimental animals, we performed this procedure in eight patients with AV node reentrant tachycardia. Preoperative, intraoperative, and postoperative electrophysiologic studies were performed in each patient. Under conditions of normothermic cardiopulmonary bypass and during atrial pacing at a constant rate with continuous monitoring of AV conduction, nine separate 3 mm cryolesions (-60 degrees C for 2 min) were placed at predetermined sites around the triangle of Koch in the lower right atrial septum. Postoperatively, each patient had a single AV node conduction curve. No patient had AV node reentrant tachycardia induced or has experienced AV node reentrant tachycardia clinically during a follow-up of up to 5 years. The cryosurgical procedure had no detrimental effects on the AH or HV interval or on the paced cycle length at which AV node Wenckebach occurred. Based on these results, this curative operation offers promise for patients with AV node reentrant tachycardia that is refractory to medical treatment.

Junctional echoes with slow retrograde conduction without His bundle depolarization: further evidence of reentry within the atrioventricular node

We report a case of intranodal reentry with slow retrograde conduction and atrial echoes in the absence of His bundle activation. Echoes were related to delay in intranodal conduction. Reentry using anomalous atrioventricular connexions is impossible without ventricular activation. This observation suggests reentry within the node without participation of neighboring structures.

Atrioventricular nodal reentrant tachycardia: studies on upper and lower 'common pathways'

Electrophysiologic studies were performed in 28 patients with documented atrioventricular (AV) nodal reentrant supraventricular tachycardia (SVT) to investigate the presence of AV nodal tissue situated between the tachycardia circuit and both the atrium (upper common pathway, UCP) and the His bundle (lower common pathway, LCP). All patients demonstrated a 1:1 AV relationship during SVT. The study protocol consisted of atrial then ventricular pacing at the SVT cycle length. UCPs were manifested in eight of 28 (29%) patients by either antegrade AV Wenckebach (six patients) or a paced atrium-His (AH) interval exceeding the AH in SVT (two patients, differences 5 and 9 msec). LCPs were manifested in 21 of 28 (75%) patients by either retrograde Wenckebach periodicity (two patients) or a paced HA interval exceeding the HA in SVT (19 patients, mean difference 25 +/- 20 msec). By these criteria, eight patients (29%) had evidence for both UCPs and LCPs. UCPs were more likely than LCPs to be manifested by Wenckebach criteria (p less than .05). Thus the AV nodal reentrant SVT circuit appears to be intranodal and is frequently surrounded by AV nodal tissue (UCP and LCP), antegrade and retrograde conduction properties of these common pathways are discordant in some cases, and conduction properties of UCP tissue differ from those of LCP tissue. These findings may have relevance in that the UCP or LCP may limit the ability of premature extrastimuli to penetrate the circuit to initiate or terminate AV nodal SVT.

The Wenckebach phenomenon

The Wenckebach phenomenon, or type I AV block, refers to a progressive lengthening of impulse conduction time, followed by a nonconducted impulse, or dropped beat. It can occur in a variety of pathologic settings, especially inferior myocardial infarction. Although a temporary pacemaker may be required, full spontaneous recovery is the rule.

Curative surgery for atrioventricular junctional ("AV nodal") reentrant tachycardia

A new surgical approach was studied prospectively in 10 consecutive patients with atrioventricular (AV) junctional reentrant tachycardia. The aim was to abolish tachycardia yet preserve normal AV conduction. On the basis of electrophysiologic study before operation, patients were classified as type A (ventriculoatrial [VA] intervals during tachycardia less than or equal to 40 ms) (seven patients) or type B (VA intervals greater than 40 ms) (three patients). Dual AV junctional pathways were demonstrable with single extrastimulus testing in seven patients before operation. Endocardial mapping during tachycardia at surgery revealed earliest atrial activation anteromedial to the AV node in type A patients and posterior to the node in the type B patients. The perinodal atrium in the region of earliest atrial activation during tachycardia was carefully disconnected from the AV node. After operation, AV junctional reentrant tachycardia was not inducible at comprehensive electrophysiologic study in any patient, and no clinical recurrences have occurred during a follow-up period of 2 to 14 months (mean 8 +/- 4). Normal AV conduction was preserved in all cases. Anterograde slow AV junctional pathway conduction was abolished in five of seven cases. Retrograde His to atrium conduction time was prolonged in type A patients but the capacity for retrograde VA conduction remained excellent. Retrograde His to atrium conduction was interrupted or severely compromised in the type B patients. These data show that there are at least two types of AV junctional reentry. Perinodal atrium appears to be part of the reentrant circuit in human AV junctional reentry. Although the most consistent effect of surgery was on the retrograde limb of the circuit, anterograde slow pathway conduction was also modified. AV junctional reentry is surgically curable with a high success rate.