The Building of the Triangular Locus of the Atrioventricular Node from Todaro to Tandler

Abstract

What is known today as the triangle of Koch (the triangular locus of the atrioventricular node, TLAVN) is bordered by the fibrous attachment of the septal cusp of tricuspid valve, the opening of the coronary sinus and tendon of valve of inferior vena cava (TIVCV). This is a concept developed cumulatively by several exceptional anatomists. The literature was reviewed with a focus on the discovery of the atrioventricular node by Sunao Tawara (January 1906), its previous announcement by Tawara’s mentor Ludwig Aschoff (1905), and the contributions of the authors who described the other components of the triangular locus. Francesco Todaro discovered the TIVCV (1865); Tawara described the atrioventricular node and its relationship with the fibrous attachment of the septal cusp of tricuspid valve and the opening of the coronary sinus. The first description of assembling all components was provided by Arthur Keith (March 1906). Keith was also the first to consider the triangular locus as a useful landmark for identifying the atrial structures of the conduction system discovered by Tawara and Wilhelm His Jr. (1893). Julius Tandler named the TLAVN as Koch’s triangle (1913). Keith’s contributions to this topic have been particularly overlooked. The “triangular locus of the atrioventricular node” or “triangle of the atrioventricular node” are more instructive and impartial names.

1. Introduction

What is known today as Koch’s triangle (the triangular locus of the atrioventricular node, TLAVN) is bordered by the fibrous attachment of the septal cusp of tricuspid valve, the opening of the coronary sinus and tendon of inferior vena cava valve (TIVCV) [1,2,3,4,5,6]. It was a construct developed by exceptional anatomists: Francesco Todaro (14 February 1839, Tripi, Messina *, Italy; 22 October 1918, Rome ^†); Sunao Tawara (5 July 1873, Nakajima, in Aki-Ōita *, Japan; 19 January 1952, Nakatsu ^†); Ludwig Aschoff (10 January 1866, Berlin *, Germany; 24 June 1942, Freiburg, Breisgau ^†); Arthur Keith (5 February 1866, Old Machar, Aberdeenshire *, Scotland; 7 January 1955, Downe, Kent ^†); Carl Eduard Walter Koch (3 May 1880, Dortmund *, Germany; 21 November 1962, Gretesch bei Osnabrück ^†); and Julius Tandler (16 February 1869, Jihlava *, Czech Republic; 25 August 1936, Moscow ^†). Selected references on the biographies of these scientists are Francesco Todaro [7]; Ludwig Aschoff [8]; Sunao Tawara [9,10,11]; Arthur Keith [12,13]; Walter Koch [3,14]; and Julius Tandler [15].

It is shown how the contributions of these authors progressed to this important macroscopic and microscopic concept—the locus of atrioventricular node. Koch’s triangle is not a fair eponym and should be replaced by a more informative and unbiased term: “atrioventricular node triangle” or “triangular locus of the atrioventricular node” (TLAVN). Previously, the late Thomas James [3,16] criticized the correctness of this eponym and, controversially, the utility of the concept [3]. Since then, the utility of the concept has been confirmed [17]. However, the criticisms of Thomas James about the soundness of the eponym, although valid, have two important gaps: (i) he assumed the contributions made by Tawara but did not describe them [18,19], and (ii) he was unaware of the Arthur Keith contributions [6]. As far as I am aware, the contributions of Arthur Keith on this topic have not been considered elsewhere, except for a brief mention in a recent paper [11].

The literature is reviewed with a focus on the original contributions reporting the components of the TLAVN [6,19,20,21,22,23,24]. It is shown that Arthur Keith was the first to assemble all components into a triangular locus [6].

2. The Announcement of the “Knoten” (Node) and Its Place

At a congress of the German Physiological Society held in Marburg, 14–16 June 1905, the “Knoten” (node) was announced for the first time [22]. Ludwig Aschoff, Tawara’s mentor and head of the Anatomical Pathology Institute at Marburg, Germany, provided a succinct but clear summary of Tawara’s ongoing investigations and presented microscopic preparations produced by Tawara [22]. A significant excerpt from Aschoff’s seminar is presented below:

“Atrial fibers first appear as a closed bundle at the coronary sinus. They then run anteriorly along the atrial septum [atrial myocardium] to the area of the insertion of the medial tricuspid leaflet [septal cusp of tricuspidal valve]. They form a tangled network and then, together with the plexus from which the fibers of the ventricular segment arise, form the so-called node. This lies as a flat, ridged structure in the fibrous septum [fibrous skeleton of heart] between the tricuspid and mitral insertions, and from it the ventricular fibers pass anteriorly through the fibrous septum [fibrous skeleton of heart] in an easily traceable bundle to the ventricular septum”.

This communication was published on September 26 in the “Münchener Medizinische Wochenschrift” (Munich Medical Weekly) [22]. Tawara himself had produced an account of his nearly completed doctoral thesis, “Die Topographie und Histologie der Brückenfasern” (The Topography and Histology of the Bridge Fibers [Penetrating part of atrioventricular bundle], which was received by the “Zentralblatt für Physiologie” (Central Journal for Physiology) on 17 April 1905 and published at the end of Jan 1906 [23]. Tawara anticipated much of the material that would later be developed in his 1906 monography [19], which now has an English translation [18]. In that paper, Tawara emphasized the following:

“So far, I have only been able to identify one region that has been precisely characterized histologically… its center is in a mesh-like node in the atrial septum [atrial myocardium] directly above the atrioventricular fibrous septum [fibrous skeleton of the heart]. From this node, a fiber bundle runs backward, often to the anterior border of the coronary sinus. A second fiber bundle runs anteriorly downward through the atrioventricular fibrous septum [fibrous skeleton of heart], where it continues with the system of Purkinje fibers [subendocardial branches]…”.

Tawara also noted that the system is present in birds and mammals. He particularly emphasized that the atrioventricular node is fully irrigated and defined its microscopic aspects:

“These fibers are characterized by various features: …they are narrower than the remaining fibers of the atrial muscle and are extraordinarily rich in nuclei and exhibit slight fibrillar differentiation, limited only to the peripheral layer. The protoplasm of these fibers always stains a distinct color than the other muscle fibers. In their imperfect development, they resemble embryonic muscle fibers…”.

In the second half of February, Tawara [21] also published a histological analysis of Hering’s (Henrich Ewald Hering, 3 May 1866, Vienna *, Austria; 16 December 1948, Papenhusen ^†) material on atrioventricular bundle transection [25] (Figure 1A). In this publication, he depicted the positions of the atrioventricular node and atrioventricular bundle, which would later be further described in his monography (Figure 1B) [19]. Previously, Hering [25] had provided macroscopic photographs illustrating in actual hearts where the atrioventricular bundle should be sectioned to produce atrioventricular dissociation (Figure 2).

3. The Development of the Concept of a Triangular Locus

James Mackenzie (12 April 1853, Pictonhill Scone *, Scotland; 25 January 1925, London ^†) read the Aschoff’s report [22] while in Burnley and communicated it to Arthur Keith, who had been struggling unsuccessfully to reproduce the results of His Jr. [24] (Wilhelm His Jr., 29 December 1863, Basel *, Swiss; 10 November 1934, Brombach-Lörrach ^†) and Hering [25] on the atrioventricular bundle. Subsequently, on 3 February 1906, Keith sent a letter to The Lancet, which was published on 3 March 1906 [6]. In it, Keith exposed that he had previously been preparing a letter to The Lancet denying the existence of the bundle:

“There may be an error in my observations or in my technique. I simply draw attention to the fact that I have failed to find a muscular bundle in the position described by His and Hering and that before it is accepted as the anatomical basis of Stokes-Adam’s disease its constancy must be proved. At least it is well to have the existence of the bundle of His [atrioventricular bundle] verified at the very commencement, so that it may not have a circulation of two centuries before it is called in as false coin. Some days after writing the part of this letter given above Dr. James Mackenzie sent me a cutting from the “Münchner Medizinische Wochenschrift” (Munich Medical Weekly) of 26 September 1905 p. 1904, in which Dr. L. Aschoff gave an account “ Ueber die Untersuchungen des Herrn Dr. Tawara, die Bruckenfasern betreffend und Demonstration der zugehorigen mikroskopischen Praparate” (On the investigations of Dr. Tawara, concerning the bridge fibers’ [penetrating portion of atrioventricular bundle] and demonstration of the associated microscopic preparations) …With this account to guide me I renewed my search with success”.

It is worth noting that Keith, although he had read Aschoff’s description of atrioventricular node [22], did not mention the description of the “Knoten” in his letter [6]. He appears to have been more focused on identifying the penetrating portion of atrioventricular bundle as he confirmed when he wrote:

“Since it is probable that others may find the initial difficulty, I experienced in dealing with this intricate region of the heart I propose to give briefly by the aid of a diagram some of the points which have proved most useful to me” (Figure 3A).

This diagram (Figure 3A) is the first description of the landmarks delimiting on the right atrial side, a triangular area (red dotted lines, Figure 3A), whose components are known today as Koch’s triangle [1,2,3,4,5,6]. Unfortunately, Keith did not portray the “Knoten” (node) described in Aschoff’s letter, showing only the atrioventricular bundle [6,22]. However, he described, as a novelty, the tendon of the inferior vena cava valve:

“In the musculature beneath the endocardium will be found two fibrous bands [see g, h, in Figure 3A] which are of great interest but, as far as I know, have never been described. These bands begin in the membranous space [membranous septum] and run through the musculature of the septum [not interatrial septum but right atrial myocardium] to the inferior vena cava valve (g) and valve of coronary sinus (h)…” [6].

Keith was unaware that Francesco Todaro [20] had described the same structure in 1865 (Figure 4). Todaro’s description [20] is more complete than that of Keith: Todaro in addition gave a microscopic description of the TIVCV: “Microscopic examination of this tendon shows it to be composed of elastic fibers and connective tissue fibers in smaller proportions”.

“I immediately noticed that from the right posterior part of the tendinous center [right fibrous trigone], a fibrous ribbon started [tendon of inferior vena cava valve]. I followed it, with great care, among the muscular fibers below and accompanied it up to the posterior pillar of the oval fossa [posterior limbus of fossa ovalis]… it entered through the posterior horn, which originates here, in the valve, following along its entire free margin, I have always found the tendon constant, even when the inferior vena cava valve is almost deficient“.

Keith also described the nodal artery: “…large artery (with a caliber of 1 to 1.5 mm) invariably perforates the right fibrous trigone just behind the bundle; the artery is always bigger than the muscular bundle at the point of perforation” [6]. Nevertheless, Keith described a component which is no longer mentioned today in the triangular locus of the atrioventricular node: a separate tendon arising from coronary sinus valve. However, it is possible that Keith was describing a case with a double tendon, as reported previously [3].

It is almost certain that Keith also read Tawara’s 1905 paper [23] because one description was not present in Aschoff’s letter [22]. In The Lancet, he wrote:

“…as Tawara described, fine, multinucleated cells with indistinct striation; I observed also, as he did, that the fibers were accompanied by numerous nerve fibers, and there are cells, too, which have the appearance of ganglionic nerve cells”.

Keith’s intention was to provide a more didactic presentation of Tawara’s discoveries. Indeed, Keith was also the first to consider the triangular locus as a useful landmark for identifying the atrial structures of the conduction system discovered by Tawara [19,23] and Wilhelm His Jr. [24,27]. He was also the first to include the tendon of the inferior vena cava in this area as a useful delimitation [6].

From the above, we see that Keith had meticulously read the reports of Aschoff and Tawara. Why did he neglect such an important and enigmatic structure as the node? The response to this question was obtained in a paper he published jointly with Martin Flack (20 March 1882 in Borden, Kent *, England; 16 August 1931 in Halton ^†) [28]. Keith in his previous report dissected, macroscopically, the bundle in a human heart [6] and now found jointly with Flack [28] (Figure 5A,B) that the node is especially difficult to see macroscopically in humans. They wrote:

“The bundle may be then seen as an easily separated strand from one to two millimeters wide-but very frequently it is so freely mixed with the tissue of the fibrous body [right fibrous trigone] that it cannot be isolated; in microscopic sections this is found to be most frequently the case. In the sheep’s heart the plexus or network is very pronounced (“Knotten” as Tawara names it) but in the human heart its features are not so characteristic. Yet in the human heart the a. v. bundle [atrioventricular bundle] has distinctly a reticulated arrangement of fibers in and near the right fibrous trigone, and it will be convenient to refer to it as the a. v. reticulum [atrioventricular node].

In Figure 5 [Figure 5B in our manuscript] the central fibrous body [right fibrous trigone] has been opened up, the muscular reticulum [atrioventricular node] removed, and the a. v. bundle shown in front of the central fibrous body passing forwards in the pars membranacea septi [Membranous part of interventricular septum]. “Indeed, this paper is a proof of the concept of the utility of the triangular locus which he previously described [6]. The position of the atrioventricular node is shown to be lying over the right fibrous trigone or at the apex of TLAVN. (Figure 5B).

Keith and Flack [28] gave a magistral macroscopic description of the fibrous skeleton of the heart and its relationship with the conduction system. They amplified and extended in this respect what was described by Tawara [19] and His [24,27].

Scientists are human beings subject to flaws and virtues. Keith has shown himself to be humble, grateful, and scientifically honest in this and on other occasions [11,12]. They wrote about the Tawara’s work:

“We take this opportunity of clearly stating that although some of our observations are new our work is in the main, but a verification of the accurate and complete monograph published recently by Tawara, a Japanese working in the laboratory of Professor Aschoff of Marburg”.

One question remains: Why was and is the delimitation of this triangular area so important? To answer this question, we need to return to one of the most important discoveries in cardiology: the discovery of the muscular bundle connecting atrial chambers to ventricular ones or the penetrating portion of the atrioventricular bundle [24,27]. At that time there was a fierce controversy about the myogenic and neurogenic theory of heart contraction [11,27]. His Jr. discovered the penetrating, non-branching and branching portions of atrioventricular bundle, as well as the right and left bundle branches. In addition, he verified that muscle contraction was present even before the formation of the nervous system. Later, he sectioned the bundle producing atrioventricular dissociation [24,27]. Although these discoveries are milestones, they were poorly divulgated and His could have been not so concise in the anatomical description of the atrioventricular bundle. His utilized microscopic cuts, not mentioning macroscopic dissection, and showed only two figures in mammals: a frontal cut in a mouse and a sagittal cut in a newborn. He wrote [24]:

“The bundle originates from the posterior wall of the right atrium, near the atrial septum, in the atrioventricular groove, attaches to the superior edge of the ventricular septum muscle with multiple fiber exchanges, travels anteriorly along it until, near the aorta, it bifurcates into a right and left branch. which latter ends in the base of the aortic cusp of the mitral valve [anterior leaflet of left atrioventricular valve]”.

We can have an idea of the difficulty in reproducing His’s results when an experienced and extremely able anatomist as Arthur Keith was unable to reproduce those findings by himself. For instance, Hering had to contact His Jr. to obtain further information on the bundle transection [11,25]. In his letter to Lancet, Keith did not mention the use of microscopic cuts. This could have added an initial additional difficulty to him before the Aschoff report [22]. On other hand Keith [6] and Keith and Flack [28] are among the pioneers in demonstrating the possibility of gross dissection of the atrioventricular conduction system, the atrioventricular bundle and its right and left branches, primarily in ungulates but also in humans. This has been confirmed by others [30,31,32]. His Jr. published in 1933 an unescapable review of the atrioventricular bundle history [27]. He wrote:

“It was ten years before the anatomists took cognizance of the A-V bundle. In 1903, Retzer, at the behest of Spalteholz, and Braeunig, at the suggestion of Engelmann, tested my results and fully substantiated them. In 1906, under the leadership of Aschoff, the important work of Tawara appeared…”,

Braeunig [33] and Retzer cited in [33] used both macroscopic and microscopic approaches. His Jr. [24] and Braeunig preparations [33] are shown in Figure 6A,B. Tawara’s merit [19,23]. was not only in confirming and expanding on His Jr.’s discoveries but also in discovering the atrioventricular node and the subendocardial branches forming a network of specialized conducting intraventricular fibers. He also discovered that these structures were connected.

Tawara’s monography “Das Reizleitungssystem des Säugetierherzens. Eine anatomisch-histologische Studie über das Atrioventrikular bündel und die Purkinjeschen Fäden” [19] (The Conduction System of the mammalian heart. An anatomical-histological study of the atrioventricular bundle and Purkinje’s fibers) [18] was completed on Dec 24, 1905, in Marburg and published at the beginning of 1906. Aschoff’s preface was written on Mar 19. In this work, Tawara further developed concepts previously announced by his mentor and himself. He described the node (“knoten”) in a human heart (No. 136. a two-year-old child’s heart). The description is provided in the section entitled “Topography of the Atrioventricular Connecting System.” Previously, Tawara had examined 10 dog hearts, in which an atrial segment of the connecting system was described. The description was compatible with that of the node. Although he drew the node in dog heart No. 120 (our Figure 6C), the term “Knoten” (node) was not mentioned in the legend.

Consequently, Tawara crystallized the concept of the node (“Knoten”) when he studied human heart No. 136 (Figure 6C,D).

“By examining the staged microscopic sections, a peculiar muscle group is observed in section No. 105, which is entirely different from the atrial musculature in terms of abundance and form of nuclei as well as in the arrangement of the muscle fibers. The muscle group is located about 1.5 mm below the lowest attachment of the noncoronary aortic leaflet, in other words, at the infero-posterior margin of the membranous septum [membranous part of interventricular septum] It adheres to the atrioventricular fibrous septum [fibrous skeleton of heart] (s), i.e., to the origin of the aorta [aortic root]. When compared with the adjacent atrial muscle fibers, the fibers are much smaller and less differentiated. The arrangement is extremely irregular, giving the appearance of a complicated glomerulation—I call this site “the node*” (k). In the following sections, this bundle gradually becomes thicker and extends anteriorly. Consequently, the tip of the bundle protrudes into the fibrous septum [skeleton fibrosum of the heart] (section No. 110). When projected to the left endocardial surface, the extension of the bundle occupies the posterior three quarters of the noncoronary aortic leaflet. The muscle fibers connect with the adjacent ordinary atrial muscle fibers (v) posteriorly and at the righthand margin of the node (Figure 1 = section No. 110) [18]”. Section 105 is the last cut at the left of the reader in Tawara’s image above (our Figure 7), and section 110 is the first cut at the right of the reader.

Tawara’s monography provided drawings that aided both the macroscopic and microscopic location of the “Knoten” (atrioventricular node) (Figure 1A,B and Figure 8A,B).

His macroscopic figures clearly show the relationship of the atrioventricular node with the septal cusp of tricuspid valve, mitral aortic leaflet, coronary sinus opening, commissure of septal and anterosuperior tricuspid leaflets, right and non-coronary leaflets, right fibrous trigone, and membranous septum (Figure 1 and Figure 6C,D). However, the inferior vena cava tendon (Todaro’s tendon) is not described or mentioned. In some of his histological figures, he sketched the tendon but made no reference to it in the legends or the text. Tawara clearly represented the tendon in a dog’s heart (Figure 6C), and possibly in a human heart (Figure 6D). In general, the upper borders of Tawara’s other figures are too close to the fibrous skeleton to allow for the visualization of the tendon at different cut levels. Nevertheless, his descriptions and figures of the node and its relationships with surrounding structures are completely original and precise. In short, Tawara was the first to describe all components of the TLAVN, except for the inferior vena cava valve tendon. Walter Koch began studying the histology of the conduction system in 1907 at Aschoff ’s laboratory [3]. His field of work was initially on the hearts’ ultimum moriens [26]. After 1907, he starts working mainly in the recently described sinuatrial node [34,35,36,37,38]. In this area he made significant contributions, improving upon the discoveries of Keith and Flack [34,35]. From 1909 to 1912, Koch published four articles on the sinuatrial node [2,35,36,37]. The drawing of the TLAVN (Figure 3B) was first published in the ultimum moriens article of 1907 [26]. In the figure first published in 1907 [26] and reproduced in 1909 [2,35], a fold of the vena cava {inferior vena cava} is shown alongside the other structures previously described by Tawara, constructing the TLAVN (Figure 3B). In one of the 1909 publications and in that of 1912, he did not represent the TLAVN [36,37]. As he admitted, advances in the atrioventricular node had been made by others. He wrote in one of his articles [36]:

“Two special muscle systems are given: the atrioventricular conduction system, specifically the Aschoff-Tawara node [atrioventricular node], and the sinuatrial node. Determining the precise location of these starting points had to be the first important task. The location of the atrioventricular node has been precisely established by the descriptions of Tawara and by the confirming studies of Mönckeberg, Fahr, Keith and Flack, Retzer, and Dewitt. In contrast, apart from the information provided by Keith and Flack, who only briefly described the location of the sinuatrial node, no further details about this muscle system are available. In my last two publications, I gave a more detailed description of the macroscopic and microscopic behavior of the sinuatrial node in humans and mammals and determined the extent of the sinuatrial node”.

Koch did not recognize the TIVCV but instead described it as a fold. It was Tandler who later suggested that this fold should be Todaro’s tendon (Figure 4) [1]. The main objective of the figure was to show the area where he found the ultimum moriens of the human heart [26] (our Figure 3B is the improved version of the original he published later [2]). In none of these papers did he mentione that Keith was the first to assemble the TLAVN components and use them as useful landmarks for finding the atrial components of the conduction system of heart. Excepting the “ultimum moriens” report [26] in all of them, the emphasis is about his findings in the sinuatrial node and the conjectures about its probable function [2,35,36,37]. He wrote [2]:

“On the other hand, at another location in the entry of the atrial cavity, more anteriorly, near the upper border of the right atrial appendage [right auricle of heart], there is a peculiar muscular system discovered by Keith and described in more detail by me in a previously published work, which in many respects, it can be considered parallel to the atrial part of the conduction system [conduction system of heart] between the atrium and the ventricle. I was able to demonstrate it at the time, at the border between the atrium and the cava [superior vena cava] in the sulcus already discussed, beginning at the upper edge of the atrial appendage [right auricle of heart]. As recent observations have shown me, it appears to extend along the sulcus, its fibers extending outward and downward almost 1 cm further than I previously stated (2 cm) (Figure 3B). It splits and disappears upward into the musculature of the cava [superior vena cava] and downward into the atrial musculature…. I consider the demonstration of this muscular system discovered by Keith to be important, since with a better understanding of its function, disturbances in it may perhaps be able to be used to explain physiological and pathological cardiac processes.”

It is fascinating to note that, even two years after Keith and Flack’s report on the sinoatrial node [38], its function as a pacemaker had not gained generalized acceptance. Indeed, Keith and Flack provided a morphological description of the sinuatrial node and suggested that it could serve as the pacemaker based on Hering’s previous physiological findings [11,38].

The term “Dreieck” (triangle) is not mentioned in Koch’s papers or in his magnum opus [2,26,35,36,37,39]. It was Tandler [1] who first named the locus as Koch’s “Dreieck” (Koch’s triangle).

“Topographically, the nodule was described almost entirely consistently by all authors, although the relationship to the aortic valve was particularly emphasized. This relationship, at least for locating the nodule, does not seem particularly well chosen, since the nodule is most easily located from the right side of the atrial septum [right atrium myocardium]. Koch relocated the nodule [atrioventricular node] into the triangle he described, outlining it as follows: the caudal border is formed by the insertion of the tricuspid valve [right atrioventricular valve]. The cranial border is formed by a fold that appears when the junction of the Eustachian and Thebesian valves is stretched to the right with pincers. The fold then runs obliquely forward and downward toward the membranous septum. In my opinion, it is the result of tension on the Todaro tendon, which has already been described elsewhere. The posterior border of this triangular field is represented by the opening of the coronary sinus. Near the apex of this triangle, as Koch correctly points out, lies the node [atrioventricular node]. We already emphasized elsewhere that Koch’s deductions, as if this field were originally a sinus area [cardiac sinus venosus], are not valid for evolutionary reasons. However, Koch’s topographical definition is certainly acceptable.”

What was Koch’s motivation for describing the TLAVN? At that time, in addition to the controversy surrounding the myogenic and neurogenic theories, another major question existed regarding the location of the cardiac pacemaker. Koch, under Aschoff’s guidance, was particularly interested in determining the location of the pacemaker. The main hypothesis that guided him in this search was that the last part of the heart to stop contracting upon death [ultimum moriens] should be the pacemaker [26]. Not being a physiologist by training, he sought to observe in the hearts of aborted human fetuses, whose lives had not been successfully resuscitated, the contraction process until the heart stopped. He basically opened the right atrium and ventricle and observed which area or region was the last to contract. According to his observations, this area was the same one where Tawara had described the atrial portion of the conduction system, including the atrioventricular node [26]. This description is also more than a year later than Keith’s letter to the Lancet [6] describing the TLAVN.

He wrote: “…following the initial observation of the dying human heart and the observation of the dying rabbit heart, the area of the coronary sinus and the lower part of the septum, i.e., precisely the place where the atrial part of the conduction bundle and the node are located, are considered the ultimum moriens of the sites of rhythmic stimuli in the mammalian heart.”

These findings reinforce Aschoff’s previous, but erroneous, theory that the cardiac pacemaker would be the node described by Tawara [11]. Seeking to better define this region, he described a fold that extended from the confluence of the coronary sinus valve and the inferior vena cava valve to the membranous septum. This fold delimited the previous structures and area described by Tawara. At that time (June 1907) [26], the sinoatrial node had been recently described (April, 1907) [38], and this structure is not indicated in the heart illustration he presents. After the publication of Keith and Flack’s article [38], his research focused on expanding Keith and Flack’s findings. He then improved the presentation of his drawing and added the location of the sinoatrial node [2] (see our Figure 3B).

The original description of the TLAVN performed by Koch, in 1907 [26] and mentioned by Tandler, reads: “this area, beginning at the periphery of the coronary vein [coronary sinus], runs the atrial portion of the conduction system, and the Tawara’s node [atrioventricular node] is also located just above the tricuspid valve insertion [septal tricuspid leaflet]. Thus, one inevitably concludes that the final segment of the coronary vein [coronary sinus] from which the conduction system originates must correspond to the sinus region as the source of automatic cardiac impulses”. Before in the same paper he had described the fold: “If one firmly stretches the lower marginal fold of the fossa ovalis [fossa ovalis of right atrium], i.e., the outgrowth formed by the confluence of the Eustachian [inferior vena calva valve] and Thebesian valves [valve of coronary sinus], by pulling with pincers, it can initially be traced as an ever-decreasing fold, and further as a grayish-white stripe that runs almost perpendicularly across the septal wall to the insertion point of the posterior tricuspid leaflet (see Figure) [this is the same as Figure 3B in our manuscript.]. If one encounters the point where the stripe meets the insertion point of the tricuspid leaflet. If a needle were to pierce the pars membranacea septi ventriculorum [atrioventricular septum and Membranous part of interventricular septum], it could be seen, especially with transmitted light from the left, that extending this fold or stripe would bisect the pars membranacea [atrioventricular septum and Membranous part of interventricular septum} and encounter the septum ventriculorum [interventricular muscular septum] at a point where, according to Tawara, the conduction bundle rides on the septum [interventricular muscular septum] and bifurcates from there.”

Before the description of the sinoatrial node, were Keith’s works on the heart, and particularly on the cardiac muscle, unknown to Koch? Certainly not, Koch cites and extensively discusses the findings of Keith [40] in his 1907 work [26]. Then, it is highly probable that Koch, like many others after him, did not realize that Keith had described the TLAVN.

Some of the claims made by Tandler [1], as it is shown so far, are clearly dubious if not incorrect. Especially when he writes that all researchers before Koch did not describe the topographical relationships of the atrioventricular node in the right atrial wall. Tawara did exhaustive microscopic work on these relationships and provided macroscopic and microscopic instructions for its location by the experimentalist [19,21,22,23]. Keith [6] and Keith and Flack [28] did this from a macroscopic point of view and, to a lesser extent, microscopic. It is important to mention that this was conducted one year before Koch’s drawing. On the other hand, Tandler was correct in mentioning that the fold described by Koch was not recognized by Koch as the tendon of the inferior vena cava valve [1]. It cannot be denied that Koch made a more didactic drawing than Keith’s original [26], and he improved it in 1909 [2]. Certainly, this aspect and the influential opinion of Tandler contribute to better divulgation of Koch’s work on the TLAVN in detriment of Keith’s pioneer work.

4. Conclusions

Francesco Todaro described the tendon of the inferior vena cava valve (1865) [20]. Wilhem His Jr. (1893) [24] discovered the atrioventricular bundle. Sunao Tawara (1905, 1906) described the atrioventricular node and its relationships with the opening of the coronary sinus, the insertion of the septal cusp of tricuspid valve, and right fibrous trigone [19,21,22,23]. He did not describe the TIVCV. Arthur Keith (1906) rediscovered the TIVCV and was the first to assemble all the components of the triangular locus [6]. He proposed them as useful marks for locating the atrial components of the conduction system described by Tawara [6]. Keith and Flack (1906) in a subsequent investigation, clearly macroscopically and microscopically identified the node at the apex of the TLAVN [28]. Walter Koch (1907) published an improved Keith’s original diagram [6] of the TLAVN one year later [26]. Koch’s description is also incomplete, describing a fold instead of the tendon of the inferior vena cava valve. None of these authors realized that they were describing a triangular area. It was Julius Tandler (1913) who first recognized the locus as a triangular area [1]. In summary, the triangular locus of the atrioventricular node was a collective construct, like many human scientific endeavors [41,42]. See

Table 1. The chronology of the discoveries related to the triangular locus of the atrioventricular node (TLAVN).

Author	Date of the Report	Achievements and Comments
Francesco Todaro	1865 [20]	He described the tendon of the inferior vena calva valve (TIVCV). The tendon was identified as a fibrous structure.
Wilhem His Jr.	1893 [24]	He found a muscular bundle crossing the fibrous skeleton of the heart connecting the right atrium and ventricles. He mentioned that the atrioventricular bundle originates from the posterior wall of the right atrium myocardium
Sunao Tawara	1905, 1906 [19,21,22,23]	He discovered the atrioventricular node. He confirmed that the atrioventricular conduction system starts near the coronary sinus opening and goes along the commissure of septal cusp of tricuspid valve crossing the fibrous skeleton of the heart. He identified the node as a connecting point between the atrial and ventricular portions of the conduction system.
Arthur Keith	1906 [6]	The first report to assemble all the components of the TLAVN. He was also the first to propose the TLAVN as a useful tool for investigating the atrial portion of the conduction system. He rediscovered the TIVCV.
Arthur Keith and Martin Flack	1906 [28]	They dissected the atrial portion of the conduction system in mammals and established its relations with the fibrous skeleton of the heart. They validated the macroscopic dissection with microscopic slides. The node was rightly positioned.
Walter Koch	1907 [26]	The TLAVN was delimited as the locus of the heart’s last moriens [last part of the heart to die or contract]. He published a drawing of the TLAVN that was reproduced several times since then. He described a fold instead of the TIVCV.
Julius Tandler	1913 [1]	He recognized the triangular nature of the TLAVN and named it Koch’s triangle. He wrongly argues that no one except Koch described the topographical relationships of the atrioventricular node in the right atrial wall.

.