Letters to the Editor

	ABSTRACT

Top Abstract Letter References Reply References

The following is an abstract of the article discussed in the subsequent letter:

 
Choi, Bum-Rak, and Guy Salama. Optical mapping of atrioventricular node reveals a conduction barrier between atrial and nodal cells. Am. J. Physiol. 274 (Heart Circ. Physiol. 43): H829-H845, 1998.The mechanisms responsible for atrioventricular (AV) delay remain unclear, in part due to the inability to map electrical activity by conventional microelectrode techniques. In this study, voltage-sensitive dyes and imaging techniques were refined to detect action potentials (APs) from the small cells comprising the AV node and to map activation from the "compact" node. Optical APs (124) were recorded from 5 × 5 mm (~0.5-mm depth) AV zones of perfused rabbit hearts stained with a voltage-sensitive dye. Signals from the node exhibited a set of three spikes; the first and third (peaks I and III) were coincident with atrial (A) and ventricular (V) electrograms, respectively. The second spike (peak II)represented the firing of midnodal (N) and/or lower nodal (NH) cell APs as indicated by their small amplitude, propagation pattern, location determined from superimposition of activation maps and histological sections of the node region, dependence on depth of focus, and insensitivity to tetrodotoxin (TTX). AV delays consisted of ₁ (49.5 ± 6.59 ms, 300-ms cycle length), the interval between peaks I and II (perhaps AN to N cells), and ₂ (57.57 ± 5.15 ms), the interval between peaks II and III (N to V cells). The conductance time across the node was 10.33 ± 3.21 ms, indicating an apparent conduction velocity (_N) of 0.162 ± 0.02 m/s (n = 9) that was insensitive to TTX. In contrast, ₁ correlated with changes in AV node delays (measured with surface electrodes) caused by changes in heart rate or perfusion with acetylcholine. The data provide the first maps of activation across the AV node and demonstrate that _N is faster than previously presumed. These findings are inconsistent with theories of decremental conduction and prove the existence of a conduction barrier between the atrium and the AV node that is an important determinant of AV node delay.

	LETTER

Top Abstract Letter References Reply References

What is the role of the AV node if the AV delay occurs before it?

To the Editor: I read with interest the article by Choi and Salama (1) published in the March 1998 issue of The American Journal of Physiology. The paper presents an intriguing account of the attempt of these authors to solve a century-old problem in cardiac electrophysiology: the mechanism of conduction through the atrioventricular (AV) node. One of the central aspects of their study is that they used fluorescent imaging techniques and a potentiometric probe to study the electrical activity in the AV region. Most important, their data suggest that much of the knowledge acquired previously by several generations of electrophysiologists and morphologists regarding the basic mechanism(s) of the AV nodal delay is incorrect. For instance, Choi and Salama (1) found no evidence of either decremental conduction or electrotonically mediated delays within the AV node. Instead, they observed an "anatomic barrier" between the atrium and the AV node. Interestingly, the authors postulate that propagation does occur across such a barrier and that the longest delay in AV conduction also occurs there. To support their conjecture, they offer evidence that the conduction velocity in the AV node is much faster than previously thought. If confirmed by other authors, such perplexing results would raise very serious doubts about the functional role of the AV node itself in the transmission of the electrical impulse from the atria to the ventricles.

Our laboratory has also endeavored to characterize the AV nodal region with the use of optical mapping and potentiometric probes (2). However, our conclusions regarding the fundamental mechanisms of AV nodal conduction (3) are in disagreement with those of Choi and Salama (1). Definite resolution of this disagreement will probably require thorough investigation with the use of many of the available electrophysiological techniques, including the newly developed three-dimensional (3-D) mapping with microelectrodes and potentiometric dyes (3).

However, there are several points in the article by Choi and Salama (1) that I feel compelled to bring to your attention. The authors present a description of a structure that they interpreted to be the AV node. However, both the location and the morphological features of this structure are in sharp disagreement with the published literature and raise doubt as to whether it is the AV node. Specifically, the following three major concerns should be addressed.

1) In general, the His bundle electrogram is considered to be the main landmark when targeting the AV node during clinical electrophysiological studies as well as in basic research. Because the His bundle is located just distal to the AV node, the His bundle electrogram is used to measure conduction delay through the AV node. Choi and Salama stated that they recorded the His bundle electrogram "at the junction of the tendon of Todaro (TT) and the crista terminalis (CT)" (legend to Fig. 1 in Ref. 1). However, this imaginable junction refers to an anatomic region at or around the opening of the coronary sinus that is far away from the bundle of His (at least 10 mm in the rabbit heart). In fact, this atrial region is thought to supply the posterior-superior atrial input to the AV node. Therefore, under no circumstances can one record a His bundle electrogram from the coronary sinus area! This raises the possibility that the authors may not have targeted the small and elusive AV node because it is located inferior to the tendon of Todaro and anterior to the proposed incorrect source of the His bundle electrogram.

2) Histological data presented by the authors in Fig. 9 (with Fig. 9C reproduced on the journal cover) were interpreted as representing the AV node. The normal heart of any mammal, including the rabbit, has only a single link between the atria and ventricles, that being the AV node. However, it is well known that the AV node itself has no direct connection with the ventricular tissue. An impulse must propagate first to the bundle of His (penetrating the bundle), and then to the right and the left bundle branches, finally reaching the Purkinje system before engaging the ventricles. To ensure a lack of communication between the upper and lower chambers of the heart, a thick collar of connective tissue is present at the AV ring. This tissue can be normally seen in any histological section that includes tissues from both atria and ventricles. One can speak of a more direct link between the atria and the ventricles only in those rare congenital cases in which propagation bypasses the AV node through an accessory pathway. Figure 9 in the article by Choi and Salama (1) and, more clearly, Fig. 9C as reproduced on the cover, present a picture of the AV nodal area that apparently lacks a connective tissue barrier. Tissues of the ventricular septum and of the structure identified by the authors as the AV node seem to be fused together, with the fibers being intermingled. Furthermore, the central fibrous body, which normally isolates the His bundle from both atrial and ventricular tissue, is clearly not present in either of these illustrations.

3) The paper is inconsistent with respect to the issue of depth of field, which is critically important for the justification of sites of origin of optical signals. First, in their abstract, the authors stated that the depth of field was ~0.5 mm. They then stated that "the depth of field of the collecting lens restricted the fluorescence measurements to a layer of cells ~100 µm from the surface" (p. H831, Ref. 1). Finally, the authors refer to the AV nodal structure that is "0.75-1.25 mm thick according to histological analysis" (p. H832, Ref. 1). This suggests that they have to record signals from a depth of at least 1.5 mm simultaneously to observe three peaks originating from atrial, nodal, and ventricular layers. I fail to see how one can explain the discrepancy between these numbers: 100 µm, 500 µm, and 1.5 mm. What were the real depth of field and sites of origin of optical signals?

These concerns raise doubts about the interpretation by the authors of their signals as being recorded from the AV nodal area. One can speculate that the optical apparatus was focused onto the apex of the triangle of Koch. This would explain the simultaneous recording of atrial, His, and ventricular optical signals but would be in disarray with the stated position of the electrode recording the His electrogram. Alternatively, on the basis of the reported histology, one might try to interpret the central structure of Fig. 9C as showing a fibrous insertion of, say, valve tissue. This, however, is difficult to reconcile with the illustrated map of electrical propagation. The authors of this most exciting and provocative study are in the best position to propose a satisfactory explanation for these and other puzzling questions.

	REFERENCES

Top Abstract Letter References Reply References

1.   Choi, B.-R., and G. Salama. Optical mapping of atrioventricular node reveals a conduction barrier between atrial and nodal cells. Am. J. Physiol. 274 (Heart Circ. Physiol. 43): H829-H845, 1998[Abstract/Free Full Text].

2.   Efimov, I. R., G. J. Fahy, Y. N. Cheng, D. R. Van Wagoner, P. J. Tchou, and T. N. Mazgalev. High-resolution fluorescent imaging of rabbit heart does not reveal a distinct atrioventricular nodal anterior input channel (fast pathway) during sinus rhythm. J. Cardiovasc. Electrophysiol. 8: 295-306, 1997[Medline].

3.   Efimov, I. R, and T. N. Mazgalev. High-resolution, three-dimensional fluorescent imaging reveals multilayer conduction pattern in the atrioventricular node. Circulation 98: 54-57, 1998[Abstract/Free Full Text].

Igor R. Efimov, Department of Cardiology Cleveland Clinic Foundation Cleveland, OH 44195

	REPLY

Top Abstract Letter References Reply References

To the Editor: We thank Dr. Igor R. Efimov for raising concerns regarding our paper (4) and appreciate the opportunity to elucidate any confusion. As pointed out by Dr. Efimov, we showed that there exists a conduction barrier that produces a "step delay" between atrionodal (AN) and nodal (N) cells that is a significant portion of total AV delay and that conduction velocity across the compact node is faster than that previously suspected from microelectrode recordings. These findings challenge "slow" (decremental) conduction across the compact node (i.e., N cells) as the only mechanism responsible for AV delay. Our study used voltage-sensitive dyes and optical mapping techniques to record, for the first time, impulse propagation across the compact node at high spatial and temporal resolution. Depth resolution was achieved by recording from different optical planes of the AV node region using a narrow depth-of-field lens.

Dr. Efimov takes a narrow point of view when he states that our data "suggest that much of the knowledge acquired previously by several generations of electrophysiologists and morphologists regarding the basic mechanism(s) of the AV nodal delay is incorrect." In doing so, he ignores equally important studies, which proposed that the AV nodal delay and Wenckebach phenomena are best explained by a step delay in addition to slow conduction (13). Such a proposition has been supported by experimental and theoretical studies from established investigators (7, 11, 13, 15, 16), but direct evidence for a step delay or its location could not be obtained because of the low spatial resolution of conventional microelectrode techniques.

Dr. Efimov misinterprets our paper when he states that we found no evidence of either decremental conduction or electrotonically mediated delays within the AV node or that we observed an "anatomic" barrier between the atrium and the AV node. Our manuscript did not deal with decremental conduction other than to say that it could not be the cause of the step delay. Slow conduction did occur, but not as slowly as previously presumed. The study did not investigate a key feature of electrotonically mediated delays at N cells, that is, the expected "notches" or discrete components of N-cell action potentials observed during an acceleration or deceleration of heart rate (2). No data were presented or implied regarding an "anatomic" barrier; the data indicated the existence of a "conduction barrier" across which AN action potentials coupled to N cells, producing an electrotonically mediated step delay at the input to the compact node. The step delay produced at this "electrical" barrier was similar during normal or retrograde propagation or for a pacing AV node (see Fig. 8 in Ref. 4), as expected for a step delay. Microelectrode studies argued for decremental and/or electrotonically mediated conduction. But a reexamination of microelectrode data shows that such conduction could be consistent with our findings. For instance, a major component of AV delay resides between AN and N cells in normal and retrograde propagation (12), and measurements of conduction velocities within the N-cell region are highly dependent on the assumed propagation pathway and, thus, can be measured as slowly as 0.02 m/s or as fast as 0.125 m/s (12). Thus our data provide new experimental support for a mechanistic explanation of AV delay that has been and continues to be under investigation, and it certainly does not raise any "doubts about the functional role of the AV node."

Dr. Efimov may disagree with our findings, but his challenge is based only on his belief that we did not record from the AV node, and, in this respect, it has no merit. Three points were made to cast doubt on the source of the optical signals: 1) the location of His bundle electrogram recordings in Fig. 1, 2) the lack of fibrous collar around the AV node in Fig. 9C (cover), and 3) the depth of field of the optical system.

With regard to the first point, the His bundle electrogram is indeed a landmark of AV node research, and its temporal characteristics with respect to atrial and ventricular electrograms are well established. Dr. Efimov asserts that the His bundle electrograms in Fig. 1 are from the coronary sinus. This is surprising given the measurement of atrial-His (AH) delay and the detection of ventricular depolarization by the His electrode. Neither would be possible from the coronary sinus. The confusion may stem from the words used to describe the location of the electrode; it might have been more helpful to describe this as "at the junction of the tendon of Todaro and a line from the crista terminalis extending anteriorly along the AV groove." It should be pointed out that Fig. 1 is the only illustration in which this issue arises and that Fig. 1 dealt with the stability of our perfused heart preparation before and after it was stained with the dye. The location of the optically mapped region was clearly shown in Fig. 2A and was used throughout the study, as stated in METHODS (p. H832, Ref. 4). In all experiments, the temporal relationship between A and H or V electrograms and the optical signals indicated that peak II of the optical signals originated from the AV node. The suggestion that peak II can be recorded from the coronary sinus is frivolous because peak II fires 49.50 ± 6.59 ms after the atrial electrogram (i.e., ₁) and 57.57 ± 5.15 ms before the ventricular electrograms (i.e., ₂) (see Table 1 in Ref. 4); such signals cannot be recorded from the coronary sinus.

In addressing the second point, Dr. Efimov proposes that the histology section shown in Fig. 9 cannot be the AV node because he believes it shows a direct link between AV nodal and ventricular cells. However, as shown in the paper (4), there is no direct electrical link between the two zones. Instead, there is a substantial delay, ₂, that represents the time needed to propagate through the His bundle, Purkinje fibers, and the interventricular septum. The origin of this delay was confirmed by severing the His bundle, which had the expected result of eliminating all ventricular action potentials (Fig. 11). The latter experiment demonstrates that the AV node is not electrically coupled to the ventricular muscle below and should eliminate any doubts that we know how to locate the His bundle. Several features of the histology allowed us to identify the AV node with confidence: 1) the full slide shows the extent and dimensions of atrial and ventricular muscle and makes us confident of their identification; 2) fat cell deposits are interposed between the atrial and ventricular tissue, as seen commonly in human (L. Nichols, personal communication) and rabbit hearts (see Fig. 12A in Ref 14); 3) the location, shape, and dimensions of the zone are correct for the rabbit AV node; and 4) the tinctorial properties of these cells are consistent with those expected for nodal cells.

Nevertheless, Dr. Efimov stated emphatically that, in Fig. 9, nodal cells intermingle with ventricular muscle. Morphologists have reported considerable individual variation in the microtopography of the tissues in the AV node region and the appearance of connective tissue surrounding the AV node (14). Few studies have analyzed longitudinal cross sections of the nodal region because of technical difficulties. In one study, the fibrous tissue separating lower nodal from ventricular cells consisted of a narrow band (~50 µm) of dark-staining connective tissue (Fig. 12B in Ref. 14). In our study, four rabbit hearts were used for histological analysis, and there was the expected variability in the appearance of the fibrous tissue. In most hearts, histology sections were sliced through different layers of cells: fibrous, nodal, connective, and back to nodal tissue. Because of the shape of the node, we rarely obtained a histology section (5 µm thick) through conductile tissue, that is, through NH, N, and His bundles continuously with no interruption by connective tissue. In Fig. 9A (and Fig. 9C as reproduced on the journal cover), the histology section does not show an intermingling of nodal and ventricular cells. Instead, the tissue adjacent to the ventricular muscle has the appearance of fibrous tissue similar to that shown by others (Fig. 12B in Ref. 14). Nodal cells at the center stained "pale pink" in the region labeled "N," and fibrous tissue stained "darker pink" on the edges between conductile and ventricular cells. This identification of connective tissue can be supported by tracking its origin to the central fibrous body, seen in deeper sections (e.g., Fig. 9A at 200 mm). We chose to show this section because of the continuity of conductile tissue, resulting in an obvious overlap of activation maps on nodal cells found on the same plane. In hindsight, it might have been better to show a histology section with an extensive collar of fibrous tissue around the node and a collage of sections from different depths to superimpose excitable nodal tissue over the isochronal maps.

Past assertions by Dr. Efimov that our signals and Fig. 9 represent bypass or accessory tracts have no merit for several reasons. 1) The existence of bypass tracts is controversial (1, 8) and at best rare in rabbits, yet our results were consistently found in all hearts (>30) tested. 2) The zone delineated by AV node optical signals was electrically isolated from the ventricular tissue, indicating that the zone could not represent an accessory tract. 3) Bypass tracts as defined by James (8) in rabbits are paths of direct input to the distal node, near the His bundle, "bypassing" the compact and lower node. The assertion by Dr. Efimov would then imply that the AV node in our paper does not include the compact or the lower node. This perplexing argument requires that, for some unknown reason, distal node (NH cells) can be detected but compact node (N cells) cannot. It is also inconsistent with the dimensions of the zone delineated by peak II, which are too large for the distal node (1, 8).

The interpretation of the origins of peak II as AV node signals relied on several lines of experiments, histology being only one aspect of the study. For instance, the lack of tetrodotoxin effect on nodal action potentials provided a strong confirmation of our interpretation (17). As far as the simultaneous analysis of histology and optical maps is concerned, we agree with Dr. Efimov that future experiments will be needed to further confirm the origins of optical signals.

Finally, with regard to the third point, the manuscript is highly consistent regarding the depth of field, contrary to the assertions made by Dr. Efimov. His difficulties arise from attempts to compare different types of measurements with each other. Our abstract described the dimensions of the recording field and the depth (not depth of field) below the surface of the preparation at which the image was taken as 5 × 5 mm and ~0.5 mm deep, respectively. That is, the focal plane of the lens was 0.5 mm below the surface of the preparation. Because of space limitations, the abstract did not fully explain the reason for adjusting the focal plane or how this was done, but the paper makes this point abundantly clear (p. H834). The second term in question, described on p. H831, indicated that the depth of field of the collecting lens was 100 µm and was correctly interpreted by Dr. Efimov. The third parameter dealt with measurements of the thickness of the AV node, which fell in the range from 0.75 to 1.25 mm (p. H832). The node is an elliptical structure, which means that the largest two-dimensional cross-sectional area of nodal tissue would be in a plane 0.375-0.65 mm below the surface of the preparation as it sections through the center of the node. To obtain maps of electrical activity in the compact node at high spatial resolution, the focal plane of the collecting lens was adjusted to maximize the ratio of peak II to peak I. This approach ensured that the focal plane of the optical apparatus was aligned with the largest possible cross section of the node to obtain the highest spatial resolution in the AV node plane of focus for accurate maps of impulse propagation. Peak II was maximized by focusing on the surface and then systematically displacing the plane of focus. At a depth of 0.5 mm below the surface and a depth of field of 100 µm, cells lying 0.4-0.6 mm from the surface were detected at higher efficiency than cells outside this region. The statement by Dr. Efimov that we had "to record signals from a depth of at least 1.5 mm simultaneously to observe three peaks originating from atrial, nodal, and ventricular layers" is perplexing. As Dr. Efimov must be aware, the larger and more densely packed ventricular cells are readily detected compared with nodal cells, even when out of the focal plane. However, cells far from the focal plane are recorded with poor spatial resolution (i.e., they are blurred) and thus appear to fire synchronously. This interpretation is in line with extensive studies from our laboratory. For example, when action potentials are recorded from the ventricular epicardium, with the focal plane on the surface and the same depth of field, cells on the focal plane are the primary contributors to the optical action potentials. Cells in layers farther than the depth of field contributed <10% to the signals such that activation spread anisotropically and was closely aligned with the orientation of the fibers on the surface. Cells >200 µm below the surface had a different orientation and yet did not modify the axis of the elliptical activation pattern (9). In AV node preparations, signals from nodal cells are considerably weaker due to their smaller 1) dimensions, 2) area of excitable membrane, and 3) upstroke compared with atrial and ventricular cells. As a result, ventricular cells from an out-of-focus plane still contribute large signals relative to AV node signals.

We sense that Dr. Efimov's concerns stem from his own studies of atrial inputs to the AV node and likewise point out several problems that cast doubts on his interpretation of optical signals from the AV node (5).

1) Efimov et al. (5) detected fluorescence signals from the AV node from a large number of diodes that they interpreted to be N-cell action potentials. The spatial and temporal resolution used in these measurements were similar to those we used, yet their diode presumably recorded N-cell action potentials that were not superimposed on A or AN action potentials, contrary to intracellular microelectrode studies. For instance, in a detailed and elegant study, Billette et al. (2) mapped the distribution of AN, N, and NH cells with the use of intracellular microelectrodes and found that action potentials from AN, N, and NH cells were commonly recorded from the same site. At any location (but unknown depth), action potentials had markedly different characteristics and activation times, with some cells activating early and others late in the same region (2). The marked superimposition of these different cell types indicates that they are either intermingled or lie at different depths within the AV node region. In either case, optical recordings from such a distribution of cells should represent the sum of their action potentials and, because of their temporal delays, appear as sequential depolarization and not action potentials that originate exclusively from N cells. Thus, in the absence of supporting evidence with microelectrodes, pharmacological tests, and/or histological analysis, it is doubtful that Efimov et al. (5) recorded AV node action potentials. On the basis of this report, it would also seem pointless to apply 3-D mapping techniques if N and NH action potentials can be recorded from the surface with no interference from other cells. On the other hand, we felt that the detection of multiple depolarizations recorded by a diode was the key to identifying the node by optical techniques. We used a simple form of 3-D analysis by displacing the focal plane to estimate the depth of the cells, which produced AV node depolarizations.

2) In rabbit hearts, the dimension of the AV node recorded by Efimov et al. (5) was 6 mm, considerably greater than that reported (~2 mm) by numerous histological and microelectrode studies (2, 14).

3) Efimov et al. (5) reported that, under sinus rhythm, atrial activation spread posterior to anterior. In perfused rabbit hearts (2), activation across the interatrial septum consistently traveled anterior to posterior, in agreement with high-resolution electrogram maps reported by McGuire et al. (10) in human hearts. The opposite direction of propagation reported by Efimov et al. (5) was not due to species differences but was most likely the result of altered conduction pathways caused by the dissection procedure.

4) In a brief article, Efimov and Mazgalev (6) reversed their previous interpretation (5) and argued that signals from the AV node must exhibit two components of depolarization coming from different depths. The first component represented transitional cells, and the second represented distal cells, much like our data (3), which we discussed with these authors at an American Heart Association meeting. Their new data (Fig. 3, left, in Ref. 6) showed a 20-ms step delay between the last "transitional" cell and the first "nodal" cell, shorter than our findings (~30 ms at 300-ms cycle length, Ref. 4) but still significant.

5) Efimov and Mazgalev estimated that their optical recordings were limited to 300-500 µm in depth (legend to Fig. 1 in Ref. 6) as based on the penetration of the excitation beam in ventricular muscle. This is a serious oversimplification of the problem because it does not take into account the depth of field of the collecting lens (at the appropriate magnification and fluorescence wavelength), the precise focal plane, or the molecular extinction coefficient of the excitation and emission light by AV nodal tissue. As stated in their introduction (p. 54, Ref. 6), "the depth of microelectrode impalement is either unknown or very difficult to verify," and data were not shown to indicate that they had resolved this technical difficulty. Thus the title (6) is not merited by the data and is inequitable to investigators who are striving to develop truly depth-resolved images of electrical activity in excitable tissues.

This response should address all the concerns raised by Dr. Efimov and eliminate any doubts that we have recorded from the AV node. The notion of a step delay is not new, but direct evidence of its occurrence is an exciting new finding. Optical mapping of the AV node is a natural and potent application of the technique, and, although it is generating more controversy than answers at this time, we expect that it will eventually make important contributions to our understanding of the basic mechanisms underlying AV node conduction.

	REFERENCES

Top Abstract Letter References Reply References

1.   Anderson, R. H., M. J. Janse, F. J. van Capelle, J. Billette, A. E. Becker, and D. Durrer. A combined morphological and electrophysiological study of the atrioventricular node of the rabbit heart. Circ. Res. 35: 909-922, 1974[Abstract/Free Full Text].

2.   Billette, J., M. J. Janse, F. J. van Capelle, R. H. Anderson, P. Touboul, and D. Durrer. Cycle-length-dependent properties of AV nodal activation in rabbit hearts. Am. J. Physiol. 231: 1129-1139, 1976.

3.   Choi, B.-R., and G. Salama. Optical mapping of atrioventricular (AV) node reveals a conduction barrier between atrial and nodal cells (Abstract). Circulation 70: 83, 1997.

4.   Choi, B.-R., and G. Salama. Optical mapping of atrioventricular node reveals a conduction barrier between atrial and nodal cells. Am. J. Physiol. 274 (Heart Circ. Physiol. 43): H829-H845, 1998.

5.   Efimov, I. R., G. J. Fahy, Y. N. Cheng, D. R. Van Wagoner, P. J. Tchou, and T. N. Mazgalev. High-resolution fluorescent mapping of rabbit heart does not reveal a distinct atrioventricular nodal anterior input channel (fast pathway) during sinus rhythm. J. Cardiovasc. Electrophysiol. 8: 295-306, 1997.

6.   Efimov, I. R., and T. N. Mazgalev. High-resolution, three-dimensional fluorescent imaging reveals multilayer conduction pattern in the atrioventricular node. Circulation 98: 54-57, 1998.

7.   Jalife, J. The sucrose gap preparation as a model of AV nodal transmission: are dual pathways necessary for reciprocation and AV nodal "echoes"? Pacing Clin. Electrophysiol. 6: 1106-1122, 1983[Medline].

8.   James, T. N. Anatomy of the cardiac conduction system in the rabbit. Circ. Res. 20: 638-648, 1967[Abstract/Free Full Text].

9.   Kanai, A., and G. Salama. Optical mapping reveals that repolarization spreads anisotropically and is guided by fiber orientation in guinea pig hearts. Circ. Res. 77: 784-802, 1995[Abstract/Free Full Text].

10.   McGuire, M. A., J. P. Bourke, M. C. Robotin, D. C. Johnson, W. Meldrum-Hanna, G. R. Nunn, J. B. Uther, and D. L. Ross. High resolution mapping of Koch's triangle using sixty electrodes in humans with atrioventricular junctional (AV nodal) reentrant tachycardia. Circulation 88: 2315-2328, 1993[Abstract/Free Full Text].

11.   Mendez, C., and G. K. Moe. Some characteristics of transmembrane potentials of AV nodal cells during propagation of premature beats. Circ. Res. 19: 993-1010, 1966[Abstract/Free Full Text].

12.   Paes de Carvalho, A., and D. F. de Almeida. Spread of activity through the atrioventricular node. Circ. Res. 8: 801-809, 1960[Abstract/Free Full Text].

13.   Rosenblueth, A. Mechanism of the Wenckebach-Luciani cycles. Am. J. Physiol. 194: 491-494, 1958.

14.   Tranum-Jensen, J. The fine structure of the atrial and atrio-ventricular (AV) junctional specialized tissues of the rabbit heart. In: The Conduction System of the Heart: Structure Function and Clinical Implications, edited by H. J. J. Wellens, K. I. Lie, and M. J. Janse. Philadelphia, PA: Lea and Febiger, 1976, chapt. 3, p. 55-81.

15.   Young, M.-L., G. S. Wolf, A. Castellanos, and H. Gelband. Application of the Rosenblueth hypothesis to assess atrioventricular nodal behavior. Am. J. Cardiol. 57: 131-134, 1986[Medline].

16.   Zipes, D. P., and C. Mendez. Action of manganese ions and tetrodotoxin on atrioventricular nodal transmembrane potentials in isolated rabbit hearts. Circ. Res. 32: 447-454, 1973[Abstract/Free Full Text].

17.   Zipes, D. P., C. Mendez, and G. Moe. Some examples of Wenckebach periodicity in cardiac tissues, with an appraisal of mechanisms. In: Frontiers of Cardiac Electrophysiology, edited by M. B. Rosenbaum. The Hague: Nijoff, 1983, p. 357-375.

Bum-Rak Choi, Guy Salama Department of Cell Biology and Physiology School of Medicine, University of Pittsburgh Pittsburgh, PA 15261