Mechanisms Underlying the Reentrant Circuit of Atrioventricular Nodal Reentrant Tachycardia in Isolated Canine Atrioventricular Nodal Preparation Using Optical Mapping
Abstract—The reentrant pathways underlying different types of atrioventricular (AV) nodal reentrant tachycardia have not yet been elucidated. This study was performed to optically map Koch’s triangle and surrounding atrial tissue in an isolated canine AV nodal preparation. Multiple preferential AV nodal input pathways were observed in all preparations (n=22) with continuous (73%, n=16) and discontinuous (27%, n=6) AV nodal function curves (AVNFCs). AV nodal echo beats (EBs) were induced in 54% (12/22) of preparations. The reentrant circuit of the slow/fast EB (36%, n=8) started as a block in fast pathway (FP) and a delay in slow pathway (SP) conduction to the compact AV node, then exited from the AV node to the FP, and rapidly returned to the SP through the atrial tissue located at the base of Koch’s triangle. The reentrant circuit of the fast/slow EB (9%, n=2) was in an opposite direction. In the slow/slow EB (9%, n=2), anterograde conduction was over the intermediate pathway (IP) and retrograde conduction was over the SP. Unidirectional conduction block occurred at the junction between the AV node and its input pathways. Conduction over the IP smoothed the transition from the FP to the SP, resulting in a continuous AVNFC. A “jump” in AH interval resulted from shifting of anterograde conduction from the FP to the SP (n=4) or abrupt conduction delay within the AV node through the FP (n=2). These findings indicate that (1) multiple AV nodal anterograde pathways exist in all normal hearts; (2) atrial tissue is involved in reentrant circuits; (3) unidirectional block occurs at the interface between the AV node and its input pathways; and (4) the IP can mask the existence of FP and SP, producing continuous AVNFCs.
Extensive evidence supports the concept that dual atrioventricular (AV) nodal conduction pathways are the basis for AV nodal reentrant echo beats (EBs) and sustained AV nodal reentrant tachycardia (AVNRT).1 2 3 However, the complete reentrant circuit has yet to be demonstrated. Data from mapping and ablation studies in AVNRT suggested that the fast and slow pathways represented different atrio-nodal connections rather than the result of functional longitudinal dissociation.4 5 In addition, at least 3 types of AVNRT have been characterized,5 and multiple AV nodal inputs may exist as well.4 The anatomic locations of the reentrant pathways related to variant types of AVNRT have not been completely determined. Furthermore, although an abrupt “jump” in AH interval is believed to represent dual AV nodal physiology, it is not clear why some patients with AVNRT have no jump and other patients without AVNRT do have a jump.6
Accordingly, this study was performed to (1) characterize the reentrant circuit in 3 types of AVNRT, (2) establish the location of atrio-nodal pathways and site of initiating unidirectional block, and (3) determine mechanisms of continuous and discontinuous AV nodal function curves (AVNFCs).
Materials and Methods
AV Nodal Preparation
The AV nodal preparations used in the present study were modified from canine atrial preparation described previously.7 Adult mongrel dogs were anesthetized. Hearts were rapidly excised and perfused through the aorta with cardioplegic solution. After the ventricle, left atrium, right atrial appendage, and sinoatrial node tissues were trimmed away, the proximal and distal right coronary artery and distal left circumflex artery were separately cannulated. The preparation was perfused with Tyrode’s solution at 37°C through the coronary cannulae at a flow rate of 20 mL/min.
Fluorescent Optical Mapping System and Data Processing
The optical mapping system was constructed as described previously.8 9 The locations of the mapped area (19.5×19.5 mm) were verified using a CCD video camera as shown in Figure 1A⇓. A bipolar electrode was used to record His bundle electrograms. Pacing was performed from the anterior limbus of the fossa ovalis near the fast pathway (FP) (Figure 1A⇓). The fluorescent action potential (AP) signals were filtered at 1000 Hz. The time of activation was determined from the maximum amplitude of the APs (APA-max). The benefits and limitations of this method are discussed in an online data supplement available at http://www.circresaha.org. The conduction velocity in the different regions was determined by measuring 3 to 5 mapping sites along the direction of the propagating wavefront.
After 30 minutes of recovery from the cardioplegic solution, the preparation was stained with 0.2 mg of di-4-ANEPPS at a concentration of ≈2 μmol/L and then was continuously perfused with a solution containing cytochalasin D at a concentration of 30 μmol/L.9 Optical data were obtained either during atrial programmed stimulation or during spontaneous junctional rhythm. After optical mapping, intracellular microelectrode recordings were obtained using techniques described previously.8
Data are expressed as mean±SEM. ANOVA was performed for the AP parameters in the Table⇓. Other statistical analyses were performed using Student’s t test. The criterion for statistical significance was P<0.05.
Characterization of AV Nodal Conduction Properties
A picture of an AV nodal preparation is shown in Figure 1A⇑. The FP was identified as the earliest retrograde atrial activation sites located near the anterior atrial septum outside Koch’s triangle (KT). The slow pathway (SP) region between the coronary sinus (CS) ostium and the tricuspid annulus was divided equally into 3 zones (Figure 1B⇑). Optical APs recorded during atrial pacing at 800 ms are shown in Figure 1C⇑. Representative optical and intracellular APs obtained from the atrium, FP, transitional zone, AV node, and SP are superimposed and displayed in Figure 1D⇑. Each optical AP represents a voltage signal summated from many cells. The optical AP duration (APD) was expected to be longer than the intracellular APD as shown in Figure 1D⇑, especially in the AV nodal and SP areas. Thus, repolarization times cannot be determined accurately from these optical APs and were measured from microelectrode recordings. In general, the configuration of the optical APs appeared to be similar to the intracellular APs except in areas with slow conduction, such as the AV node or SP.
Intracellular APs recorded in different anatomic regions (see Table⇑) were characterized into 3 groups as atrial, transitional, and AV nodal cells, consistent with the results of previous studies.10 11 APs obtained from the atrium, FP, and SP zone 1 did not have significant differences (P>0.05 in all groups), suggesting that similar types of atrial tissue surrounded KT. There were no significant differences (P>0.05 in all groups for all AP parameters) between transitional cells and SP zone 2. APs recorded in zone 3, particularly in deeper layers or near the edge of the tricuspid valve, had identical configurations (P>0.5 in all groups) compared with those recorded in the AV node, suggesting a posterior extension of the AV node and therefore an asymmetrical transitional zone surrounding the AV node. The SP was located in zones 2 and 3 and consisted of 2 types of cells, one being the transitional cell and the other being the AV nodal cell. From the FP to the transitional zone and AV node, AP amplitude, maximum diastolic potential, and maximum rate of rise at phase 0 (dV/dt) gradually decreased and resulted in a progressive delay in AV nodal conduction. The longest APD at 90% repolarization was found to be within the transitional zone rather than in the FP region.
Optical APs with double peaks12 were observed within the AV node or SP zone 3, consistent with a multilayer conduction pattern.13 14 Results from one of the experiments are shown in Figure 2A⇓. The preparation was paced at 800 ms. APs with relatively early activation were recorded in the superficial layer of the AV nodal region and correlated with the first peak of the optical AP. As the microelectrode penetrated 3 to 5 cell layers deeper (determined by the numbers of cells recorded as the microelectrode advanced), AV nodal APs with relatively late activation were obtained. Similar phenomena were also observed in SP zone 3, except that in this case, only 2 to 3 layers of penetration were required. These findings indicated that the AV node and SP were covered with a thin layer of atrial tissue, leading to the development of an early component of atrial cell activation and a late component of AV nodal cell activation. The 2 components of activation seen in the optical APs could be further separated by a premature extrastimulus (A2=200 ms) as shown in Figure 2B⇓. Optical APs with secondary derivative signals and intracellular APs recorded from the FP, transitional zone, and AV node are superimposed. As the pacing interval decreased, the 2 components of activation in the secondary derivative signals became more obvious in the AV nodal area. However, optical APs in the FP remained relatively constant with a single component, indicating that most of the cell types in the FP were the same. Similar results were observed in all 6 experiments.
To further confirm the presence of atrial tissue covering the AV nodal area, phenol was applied to the AV nodal surface to destroy the superficial (≈300 μm) layers.15 16 Optical APs obtained before and after application of phenol are shown in Figure 2C⇑. The tissue was stimulated at 800 ms. A decrease in atrial electrogram amplitude in the His tracing indicated that phenol likely destroyed the surface tissue. As the result, the early activation component in the optical AP was significantly reduced after phenol application. Phenol significantly prolonged the stimulus-to-atrium interval from 19.3±2.3 to 28.3±2.0 ms (47%, n=3; P<0.05) but only slightly delayed the AH interval from 114.3±4.7 to 118.3±4.9 ms (4%, n=3; P<0.05), suggesting that phenol mainly delayed intra-atrial conduction. As the result, the shape of AVNFCs after phenol application did not differ from those before phenol application. After applying phenol, it was difficult to obtain good uniform signals from all sites. Therefore, we were unable to analyze the conduction sequence after phenol application.
Our data are consistent the conclusion reached by others17 that the double components of AP were generated by the summation of asynchronously arriving wavefronts from the different layers. Two different optical activation maps have been constructed by using the maximum dV/dt to detect the activation time from the first or the second components.12 13 In the present study, however, we selected the APA-max as the activation time to construct a single optical map. Justifications for adapting this method are described extensively in the online data supplement.
A typical activation map is shown in Figure 3C⇓. The impulse appeared to be conducted rapidly over the atrial tissue, and it spread slowly toward the AV node through the transitional zone. As seen in intracellular recordings, conduction delay began in the transitional zone. Clearly, atrial tissues surrounding KT delivered impulses to the AV node, which strongly suggests that multiple preferential (or nondiscrete) AV nodal inputs existed. On the basis of the conduction velocities, these inputs can be arbitrarily divided into FP (192.6±17.4 cm/s; n=6), SP (40.5±5 cm/s; n=6) and intermediate pathway (IP) (78.4±10.2 cm/s; n=6), as indicated by the arrows in Figure 3C⇓ (ANOVA P<0.05 between each pathway). In the presence of an asymmetrical transitional zone, anterograde conduction over the FP reached the AV node first because a lesser amount of transitional tissue was involved, whereas impulses over the IP and SP were relatively slow because activations propagated through a greater amount of transitional tissue before they arrived at the AV node. Because the stimulation electrode was placed on the anterior limbus of the fossa ovalis, atrial tissues were activated earlier than the FP was activated. Similar activation maps were obtained in 16 of the 22 preparations. The pattern of activation sequence did not change at pacing cycle lengths of 400 to 800 ms.
Delayed activation in SP zone 3 was observed in 6 other experiments. Results from a typical experiment are shown in Figure 3D⇑-F. Although 3 similar AV nodal input pathways were observed in the activation map (Figure 3F⇑), the SP was dissociated; zone 2 showed relatively normal anterograde conduction, whereas zone 3 showed delayed activation with double-peaked AP. Not only did these cells look like AV nodal cells in the microelectrode recording as shown in Figure 2⇑ and the Table⇑, but the activation wavefront came from the AV node region, consistent with the concept of a posterior extension of AV node.18 19
Characterization of AV Nodal Reentry: Slow/Fast Type
AV nodal EBs were induced by atrial programmed stimulation in 12 preparations (slow/fast, n=8; fast/slow, n=2; and slow/slow, n=2). Among these, only one exhibited sustained reentrant tachycardia (slow/fast) of >15 minutes’ duration.
Results obtained from one of the experiments with a slow/fast EB are summarized in Figure 4⇓. The AVNFC (Figure 4A⇓) demonstrated a jump at a coupling interval of 190 ms, suggesting dual AV nodal physiology. Activation maps of A2 obtained at A1A2 intervals of 350, 200, and 190 ms are shown in Figure 4B⇓, C, and D, respectively. The activation sequence of A1 at a basic pacing cycle length of 800 ms (map not shown) was similar to that shown in Figure 3C⇑ with multiple AV nodal inputs. The activation patterns of A2 at A1A2 intervals of 210 to 400 ms were similar to the map shown in Figure 4B⇓ at 350 ms. As the coupling interval decreased, conduction over the SP gradually slowed. At a further decrease in the coupling interval from 200 ms to 190 ms, the impulse reached the refractoriness of the transitional cells, and therefore unidirectional block occurred at the junction between the FP and the AV node. Anterograde conduction then shifted from the FP to the SP, resulting in an abrupt increase (jump) in the AH interval and the initiation of a slow/fast type of EB. The corresponding optical APs and EB at an A1A2 interval of 190 ms are illustrated in Figure 4E⇓ and F. The earliest retrograde atrial activation was recorded in the FP outside KT, consistent with a slow/fast type of AV nodal reentry. The reentrant circuit was complete as shown in Figure 4F⇓. After retrograde atrial activation, the impulse entered the SP and AV node again with similar sequence but no exit. The locations of conduction block and earliest retrograde atrial activation were further identified by optical movies.
Mapping data from the slow/fast EB indicate anterograde conduction shifting from the FP to the SP, with unidirectional conduction block at the junction between the FP and the AV node, likely caused by the longest APD and refractory periods present within the transitional cells near the AV node. Initial conduction delay within the transitional cells allowed the SP sufficient recovery time, and delay in SP conduction permitted the AV node and FP enough recovery time to be activated again. Anterograde activation from the AV node to the His bundle and retrograde conduction from the AV node to the FP initiated the reentrant EB, which reentered the SP and terminated in the AV node. Atrial tissue located between the FP and SP was part of the reentrant circuit, suggesting that there was no upper common pathway.
Using the criterion of an A2H2 interval duration >50 ms for a 10-ms decrease in the A1A2 coupling interval, the abrupt jump in the AV node functional curve was observed only in 2 of the 8 experiments with slow/fast type of EBs. In the other 6 preparations, EBs with similar slow/fast reentrant circuits were induced with no clear jump. The AV nodal curve from one of these experiments is shown in Figure 5A⇓. Selected activation maps at coupling intervals of 300, 220, and 210 ms are shown in Figure 5B⇓, C, and D, respectively. The activation sequence of the EB induced at the coupling interval of 210 ms was consistent with the slow/fast type of reentry. However, anterograde conduction did not shift abruptly from the FP to the SP but moved gradually from the FP to the IP (Figure 4C⇑). With a further decrease in the A1A2 coupling interval from 220 to 210 ms, anterograde conduction further shifted from the IP to the SP with an EB but no jump. The IP served as a bridge to smoothly transfer anterograde conduction from the FP to the SP during decremental atrial stimulation, resulting in a smooth AVNFC.
In 2 preparations, discontinuous AVNFCs were observed without an inducible EB. Anterograde conduction shifted from the FP to the SP but failed to propagate retrogradely to the FP.
Characterization of AV Nodal Reentry: Fast/Slow Type
Fast/slow AV nodal reentry was seen in only 2 of 12 preparations that had inducible EBs. The activation maps obtained from both experiments at a pacing cycle length of 800 ms were similar to that shown in Figure 3F⇑ with delayed SP activation in zone 3. Results obtained from one of the experiments are shown in Figure 6⇓. The AVNFC and selected activation maps are shown in Figure 6A⇓, B, C, and D. Corresponding optical APs recorded during EB are shown in Figure 6E⇓ and F. As the pacing coupling interval decreased, delayed conduction in the AV node and FP became more obvious, as shown in Figure 6B⇓ and C. At a further decrease in A1A2 coupling interval to 220 ms, anterograde conduction markedly delayed within the AV node, resulting in prolongation of A2H2 to 576 ms. Meanwhile, retrograde conduction slowly propagated through the SP, allowing the atrial tissue located near the CS ostium to be activated again as an exit site. The AV node anterogradely conducted the subsequent EB at a relatively fast speed with a short AH interval, creating the fast/slow type of AV nodal reentry. Again, no upper common pathway was present because atrial tissue located between the FP and SP participated in the reentrant circuit.
As shown in Figure 6A⇑, a classic jump was observed in the AVNFC at an A1A2 coupling interval of 230 ms. The A2H2 interval abruptly increased by 261 ms (from 335 ms at A1A2 interval of 240 ms to 596 ms at A1A2 interval of 230 ms). However, anterograde conduction between coupling intervals of 240 to 220 ms took the same pathway as shown in the corresponding activation maps in Figure 6C⇑ and D, suggesting that the jump was caused by conduction delay within the compact AV node rather than by a shifting of the conduction over another pathway. A similar jump was also seen in the other preparation with fast/slow reentry.
Characterization of AV Nodal Reentry: Slow/Slow Type
Slow/slow AV nodal EBs were induced in 2 experiments. Both preparations exhibited delayed activation in the SP, similar to the activation map shown in Figure 3F⇑. Results obtained from one of the experiments are summarized in Figure 7⇓. AVNFC and activation maps of A2 at coupling intervals of 300, 220, and 190 ms are shown in Figure 7A⇓, B, C, and D, respectively. Optical APs and His bundle electrogram recorded at 190 ms with an EB are displayed in Figure 7E⇓ and F. As the coupling interval decreased, anterograde conduction gradually shifted from the FP to the IP, while retrograde conduction propagated slowly over the SP without exit (Figure 7C⇓). With a further decrease in the coupling interval to 190 ms, both anterograde and retrograde conduction maintained the same pathway, but additional conduction delay in the SP allowed the adjacent atrial tissue to recover and therefore to be activated again as an exit site, initiating the slow/slow AV nodal reentrant beat. The propagation shifting from the FP to the IP created a smooth AVNFC (Figure 7A⇓). The slow/slow AV nodal reentrant circuit was characterized as anterograde conduction over the IP and retrograde conduction over the SP, with the earliest atrial activation located near the CS ostium. Like the other 2 types of AV nodal reentrant EBs, no upper common pathway existed.
In another experiment, similar slow/slow EBs and sustained tachycardia were induced. The reentrant tachycardia lasted >15 minutes and was terminated by burst rapid atrial pacing. The reentrant circuits and activation sequences of the EBs and tachycardia were identical to a constant tachycardia cycle length (see additional data in the online data supplement).
At the end of the experiments, the SP was selectively interrupted by a surgical incision (slow/fast, n=4; fast/slow, n=2; and slow/slow, n=2). AV nodal reentrant EBs were no longer induced in all preparations.
Three Preferential AV Nodal Input Pathways
Our data suggest that 3 preferential AV nodal input pathways exist in all normal canine hearts regardless of the presence or absence of AV nodal EBs. These pathways represent atrio-nodal connections outside the AV node rather than longitudinal dissociation within the AV node. At a short coupling interval, unidirectional block occurs, and these functional pathways can separate into discrete reentrant pathways for anterograde and retrograde conduction. These observations are consistent with more recent data4 suggesting that multiple AV nodal input pathways might exist because anterograde AV conduction remains after ablation of both the FP and the SP.
Absence of Upper Common Pathway
The reentrant circuits for the slow/fast (counterclockwise) and fast/slow (clockwise) EBs are consistent with the previous hypothesis5 and identical to those seen in patients with similar types of AVNRT. The reentrant circuit for the slow/slow EB is different from that previously proposed.5 However, anterograde conduction actually occurs over the IP and retrograde conduction over the SP rather than over 2 separate SPs. Despite different reentrant circuits, atrial tissue surrounding KT is clearly involved in all 3 types of AV nodal reentry and suggests no upper common pathway.20 It is difficult to compare our results to those of a previous study16 under different experimental conditions. Taking into account the AV nodal blood supply, it is likely that the AV node was not totally isolated from atrial tissue in that study16 ; otherwise, it would have become ischemic or infarcted in the setting of a completely dissected KT. Thus, reentry still could have involved atrial tissue. Recent optical mapping data obtained from isolated rabbit AV nodal preparations21 also suggest that functional AV nodal pathways are located outside the compact AV node and that AV nodal reentry involves atrial and transitional cells.
Unidirectional Block in the Transitional Zone
It is generally believed that in the slow/fast type of AVNRT, unidirectional block occurs within the FP because of its longer refractory period. However, our data indicate that the longest APD is located in the transitional cells, where slow conduction and unidirectional block occurred during a normal heart rate and AVNRT.22 The combination of an asymmetrical transitional zone around the AV node with its posterior extension and 3 preferential atrio-nodal pathways provide a unique model for the development of multiple types of AVNRT.
No Direct Relationship Between AV Nodal Jump and AV Nodal Input Pathways
Our mapping data indicate the following: (1) An abrupt jump in the AH interval resulted either from shifting of anterograde conduction from the FP to the SP or from abrupt conduction delay within the AV node through the same FP. (2) Anterograde conduction over the IP smoothed the transition from the FP to the SP, producing AVNRT with a continuous AVNFC. (3) Similarly, anterograde conduction over the SP without retrograde exit can cause a discontinuous AV nodal curve with no inducible AVNRT.
AV Nodal Posterior Extension
Previous studies have provided evidence to support the concept of a posterior AV nodal extension and suggested that the posterior AV nodal extension was involved in slow pathway conduction.18 19 Our data are compatible with this conclusion and further indicate that the slow pathway existed in all normal canine hearts and that both transitional cells and posterior extension of AV nodal cells likely comprise the cellular substrate of this structure.
Limitations of the Present Study
Optical mapping is technically limited in that it cannot reconstruct the 3-dimensional structure of the AV node. The APA-max method also has several limitations. Questions related to these limitations are discussed further in the online data supplement.
The new observations in the present study have several clinical implications. First, the finding that the SP is involved in all 3 types of reentrant circuit explains the mechanism of cure by SP ablation applied in patients with various types of AVNRT. Second, multiple retrograde atrial exit sites will be expected in atypical AVNRT because retrograde activation conducts from the AV node to the posterior extension region, where multiple connections exist between the SP and atrial tissue near the CS. Finally, the FP is not involved in the reentrant circuit of slow/slow EBs, consistent with previous findings5 that FP ablation is not effective in some patients with atypical AVNRT.
Dr Jianyi Wu was supported in part by a Fellowship Award from the North American Society of Pacing and Electrophysiology. This research was also supported in part by grant 9930347Z from American Heart Association. We appreciate the generous help and advice of Dr Igor R. Efimov during the construction of our optical mapping system.
Original received December 27, 2000; revision received April 30, 2001; accepted April 30, 2001.
- © 2001 American Heart Association, Inc.
Moe GK, Preston JB, Burlington H. Physiologic evidence for a dual A-V transmission system. Circ Res. 1956;4:357–375.
Schuilenburg RM, Durrer D. Atrial echo beats in the human heart elicited by induced atrial premature beats. Circulation. 1968;37:680–693.
Patterson E, Scherlag BJ. Longitudinal dissociation within the posterior AV nodal input of the rabbit: a substrate for AV nodal reentry. Circulation. 1999;99:143–155.
Hirao K, Scherlag BJ, Poty H, Otomo K, Tondo C, Antz M, Patterson E, Jackman WM, Lazzara R. Electrophysiology of the atrio-AV nodal inputs and exits in the normal dog heart: radiofrequency ablation using an epicardial approach [see comments]. J Cardiovasc Electrophysiol. 1997;8:904–915.
Otomo K, Wang Z, Lazzara R, Jackman WM. Atrioventricular nodal reentrant tachycardia: electrophysiological characteristics of four forms and implications for the reentrant circuit. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. 3rd ed. Philadelphia, Pa: WB Saunders Co; 2000:504–521.
Tai CT, Chen SA, Chiang CE, Lee SH, Wen ZC, Chiou CW, Ueng KC, Chen YJ, Yu WC, Huang JL, Chang MS. Complex electrophysiological characteristics in atrioventricular nodal reentrant tachycardia with continuous atrioventricular node function curves. Circulation. 1997;95:2541–2547.
Kim KB, Rodefeld MD, Schuessler RB, Cox JL, Boineau JP. Relationship between local atrial fibrillation interval and refractory period in the isolated canine atrium. Circulation. 1996;94:2961–2967.
Biermann M, Rubart M, Moreno A, Wu J, Josiah-Durant A, Zipes DP. Differential effects of cytochalasin D and 2,3 butanedione monoxime on isometric twitch force and transmembrane action potential in isolated ventricular muscle: implications for optical measurements of cardiac repolarization [see comments]. J Cardiovasc Electrophysiol. 1998;9:1348–1357.
Wu J, Biermann M, Rubart M, Zipes DP. Cytochalasin D as excitation-contraction uncoupler for optically mapping action potentials in wedges of ventricular myocardium. J Cardiovasc Electrophysiol. 1998;9:1336–1347.
van Capelle FJ, Janse MJ, Varghese PJ, Freud GE, Mater C, Durrer D. Spread of excitation in the atrioventricular node of isolated rabbit hearts studied by multiple microelectrode recording. Circ Res. 1972;31:602–616.
Anderson RH, Janse MJ, van Capelle FJ, Billette J, Becker AE, Durrer D. A combined morphological and electrophysiological study of the atrioventricular node of the rabbit heart. Circ Res. 1974;35:909–922.
Efimov IR, Fahy GJ, Cheng Y, Van Wagoner DR, Tchou PJ, Mazgalev TN. High-resolution fluorescent imaging does not reveal a distinct atrioventricular nodal anterior input channel (fast pathway) in the rabbit heart during sinus rhythm. J Cardiovasc Electrophysiol. 1997;8:295–306.
Efimov IR, Mazgalev TN. High-resolution, three-dimensional fluorescent imaging reveals multilayer conduction pattern in the atrioventricular node [see comments]. Circulation. 1998;98:54–57.
Choi BR, Salama G. Optical mapping of atrioventricular node reveals a conduction barrier between atrial and nodal cells. Am J Physiol. 1998;274:H829–H845.
Barber MJ, Mueller TM, Davies BG, Zipes DP. Phenol topically applied to canine left ventricular epicardium interrupts sympathetic but not vagal afferents. Circ Res. 1984;55:532–544.
Loh P, de Bakker JM, Hocini M, Thibault B, Hauer RN, Janse MJ. Reentrant pathway during ventricular echoes is confined to the atrioventricular node: high-resolution mapping and dissection of the triangle of Koch in isolated, perfused canine hearts. Circulation. 1999;100:1346–1353.
de Bakker JM, Loh P, Hocini M, Thibault B, Janse MJ. Double component action potentials in the posterior approach to the atrioventricular node: do they reflect activation delay in the slow pathway? J Am Coll Cardiol. 1999;34:570–577.
Inoue S, Becker AE. Posterior extensions of the human compact atrioventricular node: a neglected anatomic feature of potential clinical significance. Circulation. 1998;97:188–193.
Medkour D, Becker AE, Khalife K, Billette J. Anatomic and functional characteristics of a slow posterior AV nodal pathway: role in dual-pathway physiology and reentry. Circulation. 1998;98:164–174.
McGuire MA, Lau KC, Johnson DC, Richards DA, Uther JB, Ross DL. Patients with two types of atrioventricular junctional (AV nodal) reentrant tachycardia: evidence that a common pathway of nodal tissue is not present above the reentrant circuit. Circulation. 1991;83:1232–1246.
Nikolski V, Efimov IR. Fluorescent imaging of a dual-pathway atrioventricular-nodal conduction system. Circ Res. 2001;88:e23-e30.
Mazgalev TN, Tchou PJ. Surface potentials from the region of the atrioventricular node and their relation to dual pathway electrophysiology. Circulation. 2000;101:2110–2117.