Arterial Supply to the Atrioventricular Node: Anatomical Variability and Clinical Implications

Abstract

Atrioventricular block (AVB) resulting from injury or ischaemia of the atrioventricular node (AVN) artery remains a key concern in inferior MI and procedures near the AVN. Given that ECG findings alone may be insufficient to reliably predict AVB occurrence or reversibility, anatomy-based risk assessment is essential. However, in vivo identification of the AVN artery is limited by its small size and marked anatomical variability and by the lack of specific markers for the AVN itself to confirm the artery’s terminal distribution. A main AVN artery typically arises near the cardiac crux and courses through the inferior pyramidal space, although its origin and trajectory vary. In addition, supplementary arterial supply may reach the node via anatomically distinct pathways, most commonly through the antero-inferior buttress of the atrial septum. These separate pathways variably contribute to nodal perfusion. Combined with positional variability of the AVN, this complexity limits precise anatomical risk stratification and may explain the heterogeneity of AVB occurrence and reversibility.

Atrioventricular block (AVB) is a well-recognised complication of acute inferior wall MI (inferior AMI). In a large cohort of 477 consecutive inferior AMI patients, second- or third-degree AVB developed in approximately 20% of cases.1 Additional studies have reported AVB frequencies ranging from 24% to 30%.2,3

AVB complicating AMI is often transient and may resolve in several days, particularly after reperfusion therapy. However, a significant proportion of patients have persistent block or adverse short-term outcomes, with up to approximately 14% requiring permanent pacing and having higher in-hospital mortality compared with patients without AVB.1,4,5 Current clinical decision-making relies primarily on ECG findings and clinical course, without direct evaluation of the atrioventricular node (AVN) artery. Variations in the AVN artery may influence the extent of ischaemia in the AVN, potentially affecting the reversibility of AVB.6,7

Recent reports have also highlighted the incidence of transient and permanent AVB following atrioventricular nodal re-entrant tachycardia (AVNRT) slow pathway ablation. Transient AVB occurred in 24.1% of 137 cryoablation cases and 6.4% of 204 radiofrequency ablation cases.8 Permanent AVB occurred in none of the cryoablation cases and 1.5% of the radiofrequency ablation cases.8 Although often transient, AVB remains a clinically significant complication. Notably, AVB may occur despite preservation of the AVN, likely due to swelling of the surrounding muscles.9 These findings indicate that assessment of AVB risk should consider not only the anatomical characteristics of the AVN but also those of its arterial supply.

Rare clinical reports further underscore the importance of the AVN artery. AVB has been observed in elderly patients and in those with mitral annular calcification. This occurs not only due to intrinsic conduction tissue but also as a secondary to damage, spasm or occlusion of the AVN artery.10 Coronary spasm has likewise been implicated in high-degree AVB, in some cases leading to permanent pacemaker implantation.11,12 These observations highlight the need for criteria to predict progression to permanent AVB requiring pacing.

From a clinical perspective, improving outcomes in perinodal interventions, such as AVNRT slow pathway ablation, requires a more precise anatomical risk assessment that incorporates the potential impact of arterial ischaemia. Reliable anatomical landmarks for AVN, such as the apex of Koch’s triangle, should therefore be considered in conjunction with AVN arterial anatomy.

A key anatomical feature of the main AVN artery is its course within the inferior pyramidal space (IPS), which serves as an important structural framework. This relationship has been further clarified through integrated histological and imaging-based analyses, including its spatial association with the AVN.13 These advances suggest that combined assessment ofthe IPS and the infero-septal recess may provide clinically relevant anatomical landmarks. The infero-septal recess, a recess of the left ventricular outflow tract located at the inferior aspect of the interventricular septum, has been described in prior anatomical studies.14,15

Despite these advances, understanding of the AVN artery anatomy remains incomplete. Although numerous anatomical reports on the AVN artery have been published, few have included a sufficiently large sample size to evaluate anatomical variability comprehensively.16 Even studies with more than 100 cases have primarily described the coronary origin. However, because the AVN itself cannot be directly visualised on imaging, definitive confirmation of arterial supply remains uncertain.17–19 Similarly, cadaveric studies have not directly confirmed AVN perfusion territories.20–22 Consequently, many reports classify AVN artery candidates only by their origin (right versus left), which is insufficient to capture the complexity of AVN arterial supply in vivo.

This review aims to update the anatomical understanding by presenting the actual anatomical presentation and characteristics of the AVN artery, including 3D visualisation. Although the topic addressed may be considered niche, it provides clinically relevant anatomical insight with potential implications for patient outcomes. Based on rigorous anatomical evidence, the ultimate goal of this review is to improve the resolution of the clinical grey zone surrounding AVB.

Overview of the Atrioventricular Node Artery and its Imaging Characteristics

The AVN artery typically branches from the posterolateral (PL) segment of the right coronary artery (RCA; American Heart Association [AHA] segment 4PL), near the bifurcation of the posterior descending artery, close to the cardiac crux on the posterior inferior surface of the heart.16 In left-dominant circulation, in which the posterior interventricular branch originates from the circumflex artery, the AVN artery instead arises from the left coronary artery (LCA; AHA segment 13). Importantly, however, from the standpoint of attitudinally appropriate anatomy advocated by Anderson et al., the segment is more accurately described as inferior rather than posterior.23,24 In this paper, a dual terminology has been adopted to present attitudinally appropriate terms alongside the conventional AHA terminology (e.g. PL as inferolateral and the posterior interventricular artery [PI] as the inferior interventricular artery).

The AVN artery is usually evaluated in vivo (i.e. in living subjects) using contrast-enhanced CT (Figure 1). On cross-sectional imaging, the IPS, which contains abundant adipose tissue, appears as a low-attenuation region and is readily delineated. In this space, a coronary branch near the crux ascends from the base toward the apex of the IPS. Because current imaging techniques cannot directly visualise the conduction system, vessels following this characteristic course are generally presumed to represent the AVN artery.

Figure 1: Comparison of Candidate Atrioventricular Node Arteries Identified on Contemporary Contrast-enhanced CT with Anatomically Confirmed Atrioventricular Node Arteries.

A,B: Arbitrary cross-sectional images (A: RAO 51.4°, CRA 6.1°; B: LAO 29.6°, CRA 0.0°). C–F: 3D volume-rendered images (C,E: RAO views; D,F: LAO views). The middle panels (C and D) emphasise the course within the IPS, whereas the lower panels (E and F) specifically delineate the AVN artery itself. In this case, two AVN arteries are clearly identified (arrows). In this imaging case, two candidate arterial branches are visualised (arrows); however, it cannot be definitively determined from imaging alone whether either or both represent the true AVN artery. G: Microdissection findings show that relatively large arterial branches shown in asterisks (*), which can be easily identified on imaging, do not necessarily supply the AVN but instead distribute to the interventricular muscle. In contrast, smaller branches that are difficult to identify on imaging are confirmed to supply the inferior nodal extensions of the AVN. Ao = aorta; AVN = atrioventricular node; CN = compact node; CRA = cranial; IL = inferolateral branch; INE = inferior nodal extension; IPS = inferior pyramidal space; IVC = inferior vena cava; LA = left atrium; LAD = left anterior descending artery; LAO = left anterior oblique; LCA = left coronary artery; LCX = left circumflex branch; LV = left ventricle; MS = membranous septum; NBB = non-branching bundle (His bundle); OF = oval fossa; PD = posterior descending branch; PI = posterior interventricular artery; PL = posterolateral branch; PT = pulmonary trunk; RA = right atrium; RCA = right coronary artery; RV = right ventricle; TV = tricuspid valve. Source: G: Kawashima and Sato 2018.38 Adapted with permission from Elsevier.

Several limitations, however, must be considered. As shown in microdissection findings (Figure 1G), larger-calibre branches do not necessarily supply the AVN (asterisks: interventricular branches). Conversely, the arterial supply to the inferior extensions of the AVN may arise from smaller distal branches that are not visualised on imaging. Therefore, it cannot be assumed that the AVN is located immediately distal to the artery identified on CT. Establishing reliable in vivo identification criteria and validating their clinical accuracy remain important anatomical challenges.

A recent meta-analysis of 3,919 patients assessed the AVN artery.16 Although the large sample suggests robustness, the analysis primarily reported an arterial origin: 82.41% were RCA-derived, 15.25% were LCA-derived, and 1.45% were of dual origin. Because direct identification of the AVN requires considerable expertise in anatomical dissection, most previous studies have focused on the origin of presumed AVN arteries rather than confirming their terminal distribution. Consequently, both anatomical and imaging studies have largely relied on planar assessment of branches arising near the crux and ascending linearly through the IPS.20–22 Notably, the AVN artery was reported as absent in 52.46% of anatomical cases.14 It is clear that the AVN artery is not absent in so many cases.

One role of anatomy is to provide a scientific basis for findings that are difficult to observe clinically, while also clarifying their diversity. It is therefore concerning that many studies focus solely on arterial origin, leading to limited conclusions. Beyond differences in definitions and analytical precision, true racial variation may also exist. Accordingly, the value of meta-analyses that aggregate such heterogeneous data should be carefully reconsidered.25–27

Historically, anatomical research on the AVN artery has centred on its origin, resulting in limited information about its surrounding structures and course.

Diversity in the Origin and Course of the Main AVN Arteries Branching from the Cardiac Crux Region

The AVN artery is widely recognised, but additional sources of supply exist.28,29 Therefore, it is referred to here as the ‘main’ AVN artery.

In several reports, arterial branches were described as supplying structures presumed to represent the AVN based on macroscopic anatomical criteria.30–32 Given the complex 3D architecture of the AVN, precise identification may benefit from correlation with histological findings.

Extensive microdissections of the conduction system with subsequent histological confirmation, as well as clinical anatomical analyses of the conduction system and its vascular system, have been performed (Figure 2).29,33–39 Based on these observations, macroscopic identification of the AVN was guided by: location at the apex of Koch’s triangle; typical AVN size and morphology; continuitywith the non-branching bundle (commonly referred to clinically as the His bundle) to preserve the atrioventricular conduction axis; a distinct macroscopic orientation of myocardial bundles compared with the surrounding working myocardium; and last, colour differences related to surrounding connective tissue under stereomicroscopy.35 Because these submacroscopic features may be unclear in some specimens due to preservation or other factors, inexperienced investigators should not rely solely on them for macroscopic AVN identification in research.

Figure 2: Integrated Anatomical and Histological Identification of the Atrioventricular Conduction Axis, with Histological Characterisation of Its Arterial Supply.

A–C: Macroscopic identification of the atrioventricular conduction axis. The intracardiac location of the atrioventricular conduction axis (A) and left bundle branch (LBB) (B), together with a corresponding schematic showing the sites of histological validation (C), are presented. The black dashed lines indicate the working myocardium removed to expose the conduction tissue. D–H: Histological validation corresponding to panel C. The inferior nodal extension INE (D), compact node (CN) (E), non-branching bundle (NBB) (F), branching bundle (BB) (G) and right bundle branch (RBB) (H) are identified.31 I–L: Arterial supply of the AVN in serial sections, showing: a large artery (arrow) (I) coursing through the inferior pyramidal space (IPS); branches supplying the CN (J) at the apex of the IPS; a separate branch (K) descending through the atrial septum (double arrowheads); and the peripheral AVN arteries (L) exiting toward the ventricular myocardium at the NBB.38,39 AVN = atrioventricular node; CFB = central fibrous body; CSO = coronary sinus ostium; IA = interatrial muscle; IsR = infero-septal recess; IV = interventricular muscle; MS = membranous septum; MV = mitral valve; NCC = non-coronary cusp; PL = posterolateral branch; RA = right atrium; RCC = right coronary cusp; RV = right ventricle; TMC = tricuspid-mitral continuity; TV = tricuspid valve. Sources: A–H: Kawashima and Sato 2021.35 Reproduced from Nature under a CC-BY 4.0 license. I–L: Kawashima 2025.29 Adapted from BMJ under a CC-BY 4.0 license.

Anatomical studies should designate a vessel as an AVN artery only when its distribution to the AVN is confirmed. Because precise identification of the AVN artery is not feasible based on imaging alone, it is essential to demonstrate that candidate arteries truly supply the AVN, which in turn depends on accurate macroscopic identification of the AVN itself. Therefore, Figure 2 is included to validate the reliability of the anatomical methodology for identifying the AVN, which underpins the subsequent analyses.

In this series, multiple main AVN arteries were defined as those with distinct branches to both the compact node and the inferior nodal extensions (INE; Figure 1G).38 However, in the macroscopic anatomical assessment of the main AVN artery, the INE branch shown in Figure 1G is generally difficult to discern. This may reflect variability in its anatomical expression or limitations in macroscopic identification, rather than its true absence.

The main AVN artery was identified in 101 of 103 hearts (98.1%; Table 1). In the two cases (1.9%) in which it was not observed, fine branches (outer diameter ≤0.2 mm) may have supplied the AVN.38

Table 1: Number of Main Atrioventricular Node Arteries (n=101 Hearts).

Branch	n (%)
1 branch	53 (52.5%)
2 branches	34 (33.7%)
3 branches	13 (12.9%)
4 branches	1 (1.0%)

In total, 164 main AVN arteries (mean, 1.6 per heart) were identified in 101 hearts. Based on a modified AHA coronary segmentation model, these were classified into five types (Table 2 and Figure 3).29 Here again, it should be noted that, for the sake of clinical familiarity, the conventional nomenclature of the AHA coronary segments was used; however, the recommended attitudinally appropriate terminology is also provided in parentheses thereafter.23,24

Table 2: Modified American Heart Association Classification29 of 164 Main Atrioventricular Node Arteries in 101 Hearts*.

Origin	n (%)
RCA proximal to the PI branch (segment 3) – Figure 3A	17/164 (10.4%)
Posterior descending artery (segment 4PD; inferior interventricular artery) – Figure 3B	12/164 (7.3%)
Proximal RCA PL branch (proximal 4PL segment; inferolateral branch) – Figure 3C	126/164 (76.8%)
Distal RCA PL branch (inferolateral branch) detouring around the CS (distal 4PL segment with CS detouring) – Figure 3D	3/164 (1.8%)
Distal LCX in left-dominant circulation (segment 13) – Figure 3E	6/164 (3.7%)

*Attitudinally appropriate terminology is given in the parentheses. AVN = atrioventricular node; CS = coronary sinus; LCX = left circumflex branch; PL = posterolateral; RCA = right coronary artery.

Figure 3: Anatomical Variability of the Main Atrioventricular Node Arteries.

The upper panels present the origin (left column), course and distribution (middle column), and corresponding schematic diagram (right column) of the main AVN arteries (arrowheads). The sites of origin vary as follows: segment 3 (A), segment 4PD (PI) (B), segment 4PL (IL) (C), distal segment 4PL (IL) encircling the coronary sinus (CS) (D), and segment 13 (E). Stars show the AVN. F,G: Comprehensive schematic representations of the five main AVN artery variants in the atrioventricular plane (F) and the left anterior oblique view (G). These patterns represent alternative anatomical configurations observed across different individuals and do not coexist in a single heart. 1–15 = American Heart Association coronary segments; AM = acute marginal branch; Ao = aorta; AVN = atrioventricular node; BB = branching bundle; CN = compact node; CS = coronary sinus; CSO = coronary sinus ostium; IL = inferolateral branch; IVC = inferior vena cava; LA = left atrium; LAD = left anterior descending branch; LCA = left coronary artery; LCX = left circumflex branch; MV = mitral valve; NBB = non-branching bundle; OM = obtuse marginal branch; PD = posterior descending artery; PI = posterior interventricular artery; PL = posterolateral branch; RBB = right bundle branch; RCA = right coronary artery; RV = right ventricle; TT = tendon of Todaro; TV = tricuspid valve. Sources: A–E: Kawashima and Sato 2018.39 Adapted from Elsevier under a CC-BY 4.0 license. F,G: Kawashima and Sato 2018.38 Reproduced with permission from Elsevier.

As shown in Figure 3, the AVN artery follows a complex 3D course in the IPS, rather than a planar path. For example, Figures 3A, 3D and 3E do not show a planar trajectory. Instead, a 3D trajectory is seen that is difficult to predict. Thus, regions around the coronary sinus (CS), including its left or antero-inferior aspects, cannot be assumed to be safe zones. The outer diameter at the coronary origin averaged 1.48 ± 0.52 mm (large enough to be detected on contrast-enhanced CT), whereas at the AVN distribution site it was only 0.76 ± 0.37 mm.38 This clearly demonstrates that it is difficult to visualise or trace the artery to its narrow distal portion in vivo.

In summary, a branch entering the IPS from the crux may represent the AVN artery, but not necessarily, given its variability.

Additional AVN Arteries Branching from Distinct Coronary Origins

In addition to the main AVN artery entering from the IPS near the cardiac crux, arteries with different origins and courses have been described and are sometimes referred to as atypical or anomalous AVN arteries.16,28 These include collateral pathways that may develop following occlusion of a major coronary artery.

In patients without occlusion of the main coronary trunk, however, as well as in structurally normal hearts, a descending ‘interatrial’ septal branch may supply the AVN.28,29,40

The similarities with other descending septal branches, such as Bonapace’s branch and the right superior septal artery, have been discussed previously.41,42 However, this interpretation was incorrect; these vessels should be clearly distinguished as ‘interventricular’ descending septal branches.

Meanwhile, an AVN artery arising from Kugel’s anastomosis (an anastomotic connection between the left and right coronary main trunks initially described in normal hearts) has also been reported.40,43,44 Notably, this vessel represents a subepicardial branch, and its anatomical course is fundamentally different from that of the ‘additional’ AVN artery described in our study.29

Over time, the terminology has evolved, and these vessels have sometimes been collectively referred to as ‘Kugel (collateral) artery’, particularly in the setting of coronary artery disease.45–47 In this clinical and angiographic usage, the term ‘Kugel’s artery’ potentially includes vessels with deeper or intramyocardial courses beyond the originally described subepicardial pathways.

The extent to which these deeper ‘interatrial’ septal or atrioventricular junctional vessels contribute to AVN perfusion remains incompletely understood. More recently, the presence of additional AVN arteries has been investigated.29

In hearts without coronary occlusion, macroscopic analysis confirmed additional AVN arteries reaching the AVN in 26 of 116 cases (22.4%). Thirteen cases (11.2%) involved distribution to the adjacent IPS but not the AVN itself. Histological serial sections verified AVN distribution in 5 of 14 examined cases (35.7%). Thus, at least one-third of individuals appear to possess this artery under normal conditions.

Importantly, when present, the additional AVN artery consistently formed an ‘arterial complex’ with branches supplying proximal conduction components, such as the sinus node (SN) artery or Bachmann’s bundle branch. It was also found that these arteries arise not only from the right but also from the LCA. This complex was classified into four types according to the SN artery forming the common trunk (Figure 4):29 common trunk with the right precaval SN artery, 19/26 (73.1%; (Figure 4A); common trunk with the right retrocaval SN artery, 4/26 (15.4%; Figure 4B); common trunk with the left precaval SN artery, 2/26 (7.7%; Figure 4C); and common trunk with the left retrocaval SN artery, 1/26 (3.8%; Figure 4D).

Figure 4: Anatomical Variability of the Additional Atrioventricular Node Arteries.

A–D: The additional AVN artery (red double arrowheads) consistently arises from the proximal coronary artery as a common trunk shared with the sinus node (SN) artery (orange arrowheads) and the Bachmann’s bundle branch. Owing to the marked variability in the origin of the SN artery, the additional AVN arteries are classified into four types: right precaval SN artery (A), right retrocaval SN artery (B), left precaval SN artery (C), and left retrocaval SN artery (D). Asterisks show the SN. E,F: Distally, the additional AVN artery follows a relatively constant course along the posterosuperior–anteroinferior aspect of the interatrial septum, traversing the anterior rim of the oval fossa before reaching the compact node. This course is anatomically distinct from that of the main AVN artery (red arrows), which enters the node via the inferior pyramidal space. G: A comprehensive schematic illustration of the four types of additional AVN arteries is shown in the left anterior oblique view. These arteries represent anatomically separate pathways that may variably contribute to AVN perfusion. Ao = aorta; AVN = atrioventricular node; BA = Bachmann’s bundle; BB = branching bundle; CN = compact node; CSO = coronary sinus ostium; IVC = inferior vena cava; LA = left atrium; LAA = left atrial appendage; LAD = left anterior descending artery; LCA = left coronary artery; LCX = left circumflex branch; MB = moderator band; MS = membranous septum; NBB = non-branching bundle; OF = oval fossa; PD = posterior descending artery; PI = posterior interventricular artery; PT = pulmonary trunk; RA = right atrium; RBB = right bundle branch; RCA = right coronary artery; RIPV = right inferior pulmonary vein; RSPV = right superior pulmonary vein; RV = right ventricle; SVC = superior vena cava; TV = tricuspid valve. Source: Kawashima 2025.29 Adapted from BMJ under a CC-BY 4.0 license.

Macroscopic measurements showed that the additional AVN artery had an outer diameter of 1.67 ± 0.28 mm (range 1.2–2.3 mm) at its common coronary origin. Thus, it is relatively large proximally, particularly at the common trunk near the atrial roof. Its tortuous course and marked interindividual variation suggest that it may be detectable with current imaging modalities.

Initially, the artery courses within the anterior interatrial groove. It then enters the antero-inferior buttress of the interatrial septum, through which it proceeds toward the AVN. Accordingly, the vessel narrows substantially upon reaching the septal region (0.62 ± 0.17 mm; range, 0.3–0.9 mm at entry to the compact node), with less variation than at its origin, making clinical recognition by imaging more challenging. As it descends within the septum, variation relative to surrounding structures decreases. Its course immediately anterior to the oval fossa warrants careful attention during surgical or catheter-based closure procedures.29

These findings indicate that although the main AVN artery varies considerably in calibre, it is generally large enough to be visualised by current CT technology (Figure 1). In contrast, the additional AVN artery shows limited variation and remains extremely narrow, making distal interatrial descending septal branches difficult to assess with present CT resolution.

Morphological Significance of the AVN Artery: Main and Additional Pathways with Distinct Anatomical Origins

The arterial supply to the AVN is most commonly provided by a single dominant vessel, referred to as the main AVN artery, which typically arises from the dominant coronary artery and courses through the IPS. This artery represents the principal and most consistent source of perfusion to the node.

In contrast, accumulating anatomical evidence indicates that the additional arteries may also contribute to the AVN. These arteries initially course within the anterior interatrial groove and then proceed along the anterior rim of the oval fossa. In considering their anatomical trajectory, important insights are provided by structural studies of the atrial septal musculature, which demonstrated that the posterior wall of the oval fossa is smooth rather than trabeculated, and that its superior wall does not represent a true septal structure.48 Instead, the antero-inferior buttress has been defined as the true ‘second septum’. Accordingly, the anterior rim (along which these arteries reliably run) corresponds to this antero-inferior buttress.

Given that most of the additional AVN artery courses along this ‘second septum’, its distal portion may be regarded as having septal branch characteristics.29 This supports the concept that it is fundamentally distinct from the main AVN artery in both developmental origin and anatomical significance, potentially representing a mosaic structure with subepicardial, septal and IPS-related components contributing to AVN perfusion.

Clinically, this supplementary supply may confer relative resistance of the AVN to ischaemia. However, because the additional AVN artery often shares a common trunk with the SN artery and branches to Bachmann’s bundle, forming an arterial complex, ischaemia or injury may affect multiple conduction tissues.29

Taken together, AVN perfusion is best understood as a dominant pathway through the IPS complemented by variable supplementary arteries from distinct routes, with partially overlapping functional territories depending on nodal position.

Why Anatomical Risk Mapping in the Inferior Pyramidal Space is Unreliable

Localisation of the AVN within Koch’s triangle shows substantial individual variation, but the search for a clear anatomical ‘danger zone’ and, conversely, a presumed ‘safe zone’ in the region of the AVN has long been driven by efforts to reduce the risk of ischaemic events and AVB during interventions in this area.49–53 Combined with the traditional view that the AVN artery follows a relatively straight and predictable course, these concepts suggested that careful positioning might avoid arterial injury. The present anatomical findings challenge this assumption.

Recent advances have nonetheless reinforced the apex of Koch’s triangle as a reliable anatomical marker of the AVN. In particular, a comprehensive morphological characterisation of the nodal structure within Koch’s triangle and the pyramidal space, integrated into a clinically relevant framework, represents a significant contribution.54 Furthermore, efforts to identify the apex of Koch’s triangle as an imaging landmark for estimating the precise location of the AVN have been reported.55 Collectively, these developments suggest that a foundation is being established for anatomical risk mapping aimed at avoiding AVN injury.

Detailed analysis shows that the main AVN artery typically arises from the basal region of the node and demonstrates diverse origins and courses before entering the IPS. Rather than following a single linear path, it exhibits complex 3D trajectories and expansions relative to the interatrial septum, CS ostium and septal leaflet of the tricuspid valve.38,39 Importantly, these patterns are widely distributed within the IPS.

This complexity increases with the presence of additional AVN arteries in a substantial proportion of individuals.29 These vessels arise from different sources and approach the AVN from multiple directions, further complicating the IPS. Thus, even if one arterial course can be inferred, parallel supply from another vessel extending into adjacent areas obscures presumed low-risk zones. At the same time, these additional arterial supplies suggest that the AVN may be relatively resistant to necrosis compared with regions distal to the His bundle, as described in classical anatomico-pathological studies.56,57

In this context, the AVN artery itself may represent an additional anatomical and functional marker, underscoring the need for a more detailed and systematic investigation of its course and variability.

Overall, the IPS should be regarded as a structurally heterogeneous vascular region in which AVN arterial supply may variably extend. The small calibre of these vessels and their limited visibility in vivo further complicate their identification during procedures.58,59

Accordingly, constructing a practical IPS risk map based solely on surface landmarks remains inherently challenging. This reflects not only variability in AVN location but also the complexity of its arterial supply. Thus, rather than providing precise positional guidance, anatomical analysis helps to explain the intrinsic limitations of such guidance. Recognising this constraint is essential to avoid overreliance on simplified spatial models that may create a false sense of safety near the AVN.

Implications for Perinodal Interventions

Interventions near the AVN, including ablation and right atrial device manipulation, inherently risk AVN injury.60,61 Operators have traditionally relied on landmarks such as Koch’s triangle and the CS orifice to reduce this risk.62,63 Based on the classic concept of a compact node and AVN artery located at the base of Koch’s triangle, the strategy has been to remain in presumed low-risk regions. However, the marked arterial diversity described above, in which the main AVN artery follows variable 3D courses within the IPS and additional AVN arteries approach from elsewhere, has major implications for this approach.

Because AVN arterial supply may traverse multiple overlapping pathways in the IPS, no specific IPS region or portion of Koch’s triangle can be considered a definitive safe zone.29 Areas seemingly distant from the expected course of a typical AVN artery may still receive blood from accessory vessels. Thus, proximity to classical landmarks offers only probabilistic, not absolute, information about vascular risk. Given the substantial interindividual variation and multiplicity of arterial routes, the designation of any region as completely safe is hazardous.

Nevertheless, to prevent the most serious complication, that is, AVN injury, classical landmarks, including Koch’s triangle, remain essential references during intervention.64–66 In practice, anatomical knowledge should be integrated with functional indicators such as local electrograms, junctional responses and dynamic atrioventricular conduction changes. A cautious strategy, including stepwise energy titration and prompt recognition of conduction disturbances, is crucial to minimise AVB risk. Future advances in imaging and mapping may eventually enable individualised evaluation of nodal vascular anatomy.

Implications for Inferior Myocardial Infarction

AVB is a well-known complication of inferior MI. However, its incidence, severity and reversibility can vary considerably, even between patients with seemingly similar coronary occlusion patterns.67,68 Although ischaemia of the AVN artery has long been considered the primary mechanism, recent findings regarding the anatomical variability of this vessel offer new explanations for the diversity of clinical symptoms.29,38,69,70

When an infarction occurs in the territory supplied by the main AVN artery, and the contribution of collateral arteries is minimal, ischaemia of the conduction tissue can become more severe and prolonged. This increases the risk of developing a high-degree or persistent AVB.70 In contrast, patients with dual or more distributed arterial supply may experience milder and shorter conduction disturbances due to partial perfusion, allowing recovery after ischaemia resolves.70,71 This arterial diversity provides a plausible anatomical basis for the inconsistent association between the extent of inferior wall infarction visible on angiography and the occurrence of AVB.

Furthermore, due to the small calibre and variable course of the AVN artery, the presence or absence of lesions cannot be reliably inferred based solely on epicardial angiographic findings.29,38 Even when causative lesions appear similar between patients, the affected downstream microvascular regions may differ significantly.72 This divergence may explain why clinical outcomes cannot be predicted based solely on macroscopic coronary anatomy.

Thus, recognition of the variability in AVN arterial supply improves understanding of AVB pathophysiology in inferior infarction. AVB should be viewed not as a uniform consequence of specific coronary lesions, but as the result of interaction between epicardial occlusion and highly individualised microvascular anatomy, accounting for its unpredictability and variable reversibility.

The Role of Anatomy: Defining the Limits of Prediction

Creating a reliable anatomical risk map to prevent injury to the AVN artery is difficult. At first, this may seem to reduce the clinical usefulness of anatomical studies. In reality, however, anatomy plays a crucial role in defining the limits of predictability. By clarifying the extent and nature of structural variation, it can be demonstrated that predictions based solely on location are impossible. Recent studies have shown that AVN arteries do not consistently exhibit patterns of origin, course or contribution from accessory vessels. Rather, they are widely distributed in the IPS, forming a heterogeneous arterial complex.29,38,39 This vascular network does not follow a single, linear course, and its relationship to common anatomical landmarks varies. These findings do not suggest anatomical inadequacy but rather provide essential information that illustrates the limitations of spatial inference. While anatomy indicates areas where critical structures may exist, when combined with physiological assessment, it enables clinical judgement based on probabilistic risk understanding.73 This perspective reinforces the importance of anatomical research in modern cardiology and electrophysiology.

Future Perspectives

Current knowledge emphasises the anatomical description of the AVN artery as a single linear branch originating from either the LCA or RCA.16,31 However, the actual anatomy is far more complex (Figure 5).29,38,39 Advances in recent research methodologies have deepened our understanding of the anatomically and functionally complex region surrounding the AVN. Next-generation imaging techniques, such as high-resolution cardiac CT and MRI, have progressively improved the ability to visualise myocardial microvasculature. However, routine identification of vasculature related to the conduction system, including the AVN artery, remains challenging.58,74 Due to their small calibre and significant interindividual variation in course, detailed anatomy of the AVN arteries has primarily been elucidated through anatomical studies. Once the AVN artery enters the atrial septum, it consistently descends from posterosuperior to anteroinferior (Figure 5, upper).29,34,38,67,71 After passing through the AVN, only a few of these arteries continue distally within the conduction axis. The vast majority, however, pass through the AVN and proceed directly toward the ventricle. It is hoped that future advances in clinical imaging technology will enable visualisation of the distal course of these arteries in vivo (Figure 5, lower). Integration of anatomical data with clinical electrophysiological findings, that is, the correlation of vascular patterns with intraoperative responses, conduction behaviour and clinical outcomes, suggests that indirect risk assessment is feasible even without direct vascular visualisation.71,75

Figure 5: Experimental 3D Visualisation of the Cardiac Conduction System and Arterial Supply to the Atrioventricular Nod.

A and B: The origin, course, and distribution of representative arterial pathways to the AVN shown in right anterior oblique (A) and left anterior oblique (B) views. The main AVN artery arises from segment 4PL (IL) and courses through the inferior pyramidal space, whereas an additional arterial branch originates from a common trunk of the proximal right coronary artery and reaches the node via the antero-inferior buttress of the atrial septum. These two arterial pathways are anatomically distinct in their origin and course. C–F: The intranodal course and distribution of these arteries, from the distal AVN to the proximal left bundle branch (LBB), are shown using 3D histological reconstruction from serial sections.29,38,39 C,D: Main AVN artery only. E and F: Main and additional AVN arteries. Most AVN arteries penetrate and supply the AVN, and subsequently extend beyond the node to supply adjacent myocardial structures (asterisks). These findings illustrate the anatomical variability of arterial supply to the AVN, comprising a dominant pathway through the inferior pyramidal space and variable supplementary inputs via anatomically separate routes. AM = acute marginal branch; AV = atrioventricular; AVN = atrioventricular node; BA = Bachmann’s bundle; BB = branching bundle; CB = conus branch; CN = compact node; IL = inferolateral branch; LBB = left bundle branch; NBB = non-branching bundle; OF = oval fossa; PD = posterior descending artery; PI = posterior interventricular artery; PL = posterolateral branch; RBB = right bundle branch; SN = sinus node. Sources: A,B: Kawashima 2025.29 Adapted from BMJ under a CC-BY 4.0 license.

Furthermore, even among patients with seemingly similar coronary artery lesions, the degree and distribution of downstream microcirculatory impairment can vary significantly.76 This variation is thought to contribute to differences in the severity and reversibility of conduction disturbances.72,74 In AVB complicating inferior wall MI, the presence or absence of AVN artery ischaemia and collateral circulation is also known to influence clinical outcomes.70 Hence, strategies emphasising the quantification of risk in the context of vascular variability, rather than the identification of clear anatomical safe zones, and the early detection of conduction disturbances are very likely to become crucial.

Conclusion

The arterial supply to the AVN is more complex than a single-vessel model but remains structurally organised. A dominant AVN artery, typically coursing through the IPS, provides the principal source of perfusion in most individuals, although its origin and course vary. In addition, anatomically distinct supplementary arterial pathways may reach the node via the antero-inferior buttress of the atrial septum.

These pathways are separate in their origin and course, yet their relative contribution may vary depending on the precise anatomical location of the AVN. This combined arterial and positional variability limits the definition of a consistent anatomical risk zone and underlies the clinical heterogeneity of AVB. Accordingly, although anatomical landmarks remain important, integration of anatomical understanding with functional monitoring is essential for procedural safety.

Clinical Perspective

• The main atrioventricular node (AVN) artery follows a complex, 3D, and highly variable course in the inferior pyramidal space (IPS), often involving tortuous pathways and multiple arterial contributors. Consequently, anatomically defined safe zones, including regions around the coronary sinus, are unreliable during procedures near the AVN.

• Current imaging techniques, including contrast-enhanced CT, can identify candidate proximal AVN arteries but are insufficient for delineating distal or additional AVN branches. Therefore, real-time functional assessment, such as continuous electrophysiological monitoring and stepwise energy delivery, is essential to detect early conduction changes and prevent irreversible atrioventricular block (AVB).

• The presence of additional AVN arteries contributes to both the vulnerability and the resilience of AVN conduction. This variability may underlie the heterogeneous occurrence and reversibility of AVB, supporting a cautious, stepwise procedural approach with emphasis on the early detection of conduction disturbances.

Funding Statement

This work was supported by JSPS KAKENHI (grant number: 23K11939, 26K15721).