Relevance of Conduction Disorders in Bachmann’s Bundle During Sinus Rhythm in Humans
Abstract
Background—
Bachmann’s bundle (BB) is considered to be the main route of interatrial conduction and to play a role in development of atrial fibrillation (AF). The goals of this study are to characterize the presence of conduction disorders in BB during sinus rhythm and to study their relation with AF.
Methods and Results—
High-resolution epicardial mapping (192 unipolar electrodes, interelectrode distance: 2 mm) of sinus rhythm was performed in 185 patients during coronary artery bypass surgery of whom 13 had a history of paroxysmal AF. Continuous rhythm monitoring was used to detect postoperative AF during the first 5 postoperative days. In 67% of the patients, BB was activated from right to left; in the remaining patients from right and middle (21%), right, central, and left (8%), or central (4%) site. Mean effective conduction velocity was 89 cm/s. Conduction block was present in most patients (75%; median 1.1%, range 0–12.8) and was higher in patients with paroxysmal AF compared with patients without a history of AF (3.2% versus 0.9%; P=0.03). A high amount of conduction block (>4%) was associated with de novo postoperative AF (P=0.02). Longitudinal lines of conduction block >10 mm were also associated with postoperative AF (P=0.04).
Conclusions—
BB may be activated through multiple directions, but the predominant route of conduction is from right to left. Conduction velocity across BB is around 90 cm/s. Conduction is blocked in both longitudinal and transverse direction in the majority of patients. Conduction disorders, particularly long lines of longitudinal conduction block, are more pronounced in patients with AF episodes.
WHAT IS KNOWN
Bachmann’s bundle is an anatomical structure between the right and left atrium and is considered to be the main route of inter-atrial conduction with specialized conduction properties.
Clinical and experimental studies suggested Bachmann’s bundle plays a role in development of atrial fibrillation.
WHAT THE STUDY ADDS
Bachmann’s bundle is not only activated from the right side during sinus rhythm, but also from its mid portion (septum) and left side. Moreover, the effective conduction velocity was only approximately 90 cm/s.
Conduction disorders across Bachmann’s bundle was associated with paroxysmal and early-postoperative atrial fibrillation, suggesting a possible role in AF risk.
About a century ago, Jean George Bachmann examined conduction across a muscular band of parallel, longitudinally orientated muscle fibers running from the right auricle at the superior cavo-atrial junction over the roof of the left atrium (LA) toward the left atrial appendage (LAA). This bundle, which came to be called Bachmann’s bundle (BB),1 is considered to be the preferential route of interatrial conduction. Whether this is because of the presence of specialized conduction tissue or the parallel-aligned orientation of the muscle bundles remains controversial.2–8 In vivo measurements of interatrial conduction velocity in canine hearts demonstrated that the effective conduction velocity across BB is considerably higher compared with other atrial sites.5,6,9 Creation of a surgical lesion across BB resulted in interatrial conduction block and caused biphasic P waves on the surface ECG.10 Clinical studies have demonstrated that biphasic P waves predispose to development of atrial fibrillation (AF).11–13 It was therefore assumed that conduction disorders within BB play a major role in the pathophysiology of AF, although the exact mechanism is not understood.11,14,15 To date, conduction properties of BB in humans have never been investigated in detail. In this study, we therefore performed direct high-resolution mapping of BB during sinus rhythm (SR) in patients undergoing coronary artery bypass surgery to examine (1) whether conduction disorders are present at BB, (2) the extensiveness of conduction disorders and their impact on LA excitation, and (3) differences in characteristics of conduction disorders between patients with and without AF episodes.
Methods
This study is part of a prospective observational project, entitled Quest for Arrhythmogenic Substrate of Atrial fibRillation (QUASAR), which was approved by the Medical Ethical Committee in the Erasmus Medical Center (MEC 2010–054). The QUASAR project adheres to the declaration of Helsinki principles, and written informed consent was obtained from all patients before the surgical procedure.
Study Population
Epicardial mapping was performed in 185 patients undergoing elective coronary artery bypass surgery. Patients with paced atrial rhythm, Wolff–Parkinson–White syndrome, renal failure, previous open-chest cardiac surgery, prior ablative therapy, presence of assist devices, and prior radiation for chest malignancies were excluded. Patient characteristics are summarized in Table. Thirteen patients (aged 70±5 years, 62% male) had paroxysmal AF (PAF) since 2 years (range 4 months to 23 years); the remaining 172 patients (aged 65±9years, 155 (85%) male) had no history of AF. None of the patients had a typical biphasic P wave in the inferior leads of the surface ECG. Mapping was performed in patients with a mean heart rate of 71±13 beats per minute.
| No AF | Paroxysmal AF | P Value | |
|---|---|---|---|
| No of patients (N) | 172 | 13 | |
| Age, y±SD | 65±9 | 70±5 | 0.05 |
| Male sex, % | 147 (85) | 8 (62) | 0.04 |
| BMI, kg/m2±SD | 28±5 | 28±4 | 1.0 |
| Hypertension, % | 104 (60) | 10 (77) | 0.38 |
| Hypercholesterolemia, % | 74 (43) | 5 (38) | 1.0 |
| Diabetes mellitus, % | 60 (35) | 5 (38) | 0.77 |
| Peripheral vascular disease, % | 22 (13) | 1 (8) | 1.0 |
| Thyroid disorder, % | 6 (3) | 0 | 1.0 |
| Echocardiography | |||
| LVF, % | 167 (97) | 13 (100) | 0.06 |
| Normal function, % | 130 (78) | 8 (62) | |
| Mild dysfunction, % | 30 (18) | 4 (31) | |
| Moderate dysfunction, % | 6 (4) | 1 (8) | |
| Severe dysfunction, % | 1 (1) | 0 (0) | |
| LA size, % | |||
| Dilated LA (>45 mm), % | 23 (14) | 3 (23) | 0.38 |
Epicardial High-Resolution Mapping
Epicardial high-resolution mapping was performed after sternotomy during normothermia and before extracorporal circulation. A bipolar pacemaker wire serving as a temporal reference electrode was placed at the right atrial free wal,l and a steal wire was attached to subcutaneous tissue in the thorax as an indifferent electrode. BB was mapped with electrode arrays containing either 128 or 192 unipolar electrodes (interelectrode distances: 2.0 mm) with lengths of, respectively, 32 and 48 mm; the width of both arrays was 16 mm.
Mapping of BB was performed by positioning the mapping array within the sinus transversus, behind the aorta with its tip against the LAA, as demonstrated in Figure 1. In case of the 128-electrode mapping array, the device was pulled backwards over the roof of the LA toward the superior cavo-atrial junction.
Figure 1. Mapping of Bachmann’s Bundle (BB). Top, Postmortem human heart from a right lateral superior view. Epicardial mapping of Bachmann’s bundle is performed by introducing the mapping array behind the aorta and positioning its tip against the border of the left atrial appendage (white arrow). Middle, 192 unipolar electrode mapping array (interelectrodes distances of 2 mm) covering an area of 48×16 mm. Bottom, Schematic view of the position of the mapping array on BB. The mapping area was subdivided into 3 different quadrants of 16×16 mm and labeled as right, central, and left side. AO indicates aorta; LAA, left atrial appendage; RA, right atrium; RAA, right atrial appendage; RSPV, right superior pulmonary vein; SVC, superior caval vein; and VCS, vena cava superior.
Five seconds of SR was recorded at every mapping site. The recordings included surface ECG lead I, the right atrial bipolar reference electrogram, a calibration signal with an amplitude of 1 mV, and a duration of 1000 ms. Recordings were made with a custom-made mapping system with an amplifier (gain 1000), filter (bandwidth 0.5–400 Hz), and an analogue-to-digital data converter (16 bits). All data were sampled at 1 kHz and stored on hard disk.
Analysis of the Mapping Data
Signals were analyzed with custom-made software, as previously described in detail.16–19 Color-coded activation maps of every SR beat were automatically created by marking the steepest negative deflection of extracellular potentials. An averaged SR activation map was then constructed by time alignment of all individual beats recorded during 5 seconds of SR, thereby excluding aberrant and atrial premature beats. These averaged activation maps were used for analysis of voltages, conduction velocities, conduction blocks, and patterns of activation. Voltage maps were constructed by measuring peak-to-peak amplitudes of unipolar SR potentials.
As demonstrated in the lower panel of Figure 1, the mapping array was divided into 3 equally sized quadrants (16×16mm) to examine differences in conduction velocity over the right, central, and left part of BB. Conduction velocity across BB was measured by automatically constructing isochrones at every 5 ms. The main trajectories of propagation were created perpendicular to the isochrones.18,20 For the first part, the main trajectory was constructed from the initial isochrone at 5 ms back to the onset of the wavefront. If the onset of the wavefront consisted of multiple electrodes, the electrode which resulted in the most perpendicular trajectory in relation to the isochrone was chosen as start of the wavefront. From the first isochrone, the trajectory was constructed between consecutive isochrones, choosing the most perpendicular segment to the next isochrone until the last activated electrode was reached. When the last activated site covers more than one electrode, again the electrode resulting in the most perpendicular line was chosen. Conduction velocity was subsequently calculated by summating the lengths of all segments between the isochrones and dividing it by the time difference of the earliest and latest activated electrode.
Differences in activation time (local conduction delay) between neighboring electrodes were calculated in areas of 2×2 electrodes. The maximum local conduction delay between 2 adjacent electrodes was calculated to determine the incidence of slowing of conduction and conduction block. Slowing of conduction was defined as a local conduction delay of ≥7 ms corresponding to a conduction velocity <28 cm/s and conduction block as a local conduction delay of ≥12 ms corresponding to a conduction velocity <18 cm/s, as previously described.17,18 Lengths of all lines of conduction block were measured, and they were subdivided into longitudinal or transversal lines of conduction block. Longitudinal conduction block was determined as lines of conduction block that interrupt wavefronts emerging in longitudinal direction and transverse block vice versa.
To study variation in patterns of activation of BB during SR, entry sites of SR wavefronts into BB were labeled as right atrial, central, LA, or combined entry sites. A right atrial entry site was defined as a wavefront first entering the mapping array from the right side and propagating toward the left side of BB, whereas in case of a LA entry site, activation started from the tip of the electrode positioned at the border of the LAA and spread toward the right side of BB. A wavefront emerging in the middle of the mapping array propagating to either the right and left side was labeled as a central entry site. Simultaneously activated areas (conduction velocities >200 cm/s) within BB were also labeled as central entry sites.
Postoperative Atrial Fibrillation
Early postoperative AF (PoAF) was defined as sustained AF episodes lasting >30 s. The incidence of PoAF was determined using continuous rhythm monitoring up to the first 5 days after cardiac surgery. The occurrence of PoAF was correlated with the amount of conduction block and the length of lines of the conduction block.
Statistical Analysis
Normally distributed continuous variables are presented as mean±SD and skewed data as median (minimum−maximum). Categorical data are expressed as numbers and percentages. Data were compared using either Student t test, Mann–Whitney U test, χ2, or Fisher exact test when appropriate. Correlation between voltage or conduction velocity and patient characteristics was made by using linear Pearson regression model. Adjustments were made for sex, age, body mass index, hypertension, diabetes mellitus, hyperlipidemia, peripheral vascular disease, left ventricular function, and LA dilatation. Multivariate logistic regression models were used to test the relation between slow conduction or conduction block and the same patient characteristics. As a result of the small groups with a high amount of slow conduction and conduction block, univariate analyses were done to select the determinant of interest. Age, sex, a history of AF, hypertension, diabetes mellitus, left ventricular function, and LA dilatation were chosen. In addition, the association between conduction block and PoAF was also investigated with logistic regression models. Because of the limited number of patients with PoAF, univariate analyses were performed to select properties of interest for multivariate analysis for the prediction of PoAF with conduction block. Type of conduction block, age, sex, hypertension, and LA dilatation were included in the multivariate analysis for PoAF. A P value <0.05 was considered as statistically significant.
Results
Patterns of Activation
Figure 2 shows color-coded SR activation maps of BB demonstrating the different types of patterns of activation observed in our study population. Arrows indicate main direction of propagation. In the majority of the patients (N=124, 67%), BB was activated by a single wavefront propagating from the right to the left side of BB as demonstrated in the upper left map of Figure 2. However, in 53 patients (29%), BB was activated by multiple wavefronts entering BB from different sites. In case of multiple different entry sites, wavefronts either collided or were separated by areas of conduction block. The upper right map of Figure 2 shows wavefronts entering on the right side and central part of BB, which was observed in 21% of the patients. Activation of BB from the right, left side, and central part of BB occurred in 8% of the patients. A typical example is given in the lower left map; wavefronts not only enter BB from the right and the left side, but a large wavefront also emerges in the central part of BB (dashed asterisk) and activates a large area more or less simultaneously. In the remaining 8 patients (4%), a wavefront entered in the central part of BB and spread subsequently to both the right and left side of BB (lower right map). There was no difference in incidences of the various patterns of activation between patients without a history of AF and with PAF (P=0.72).
Figure 2. Activation patterns at Bachmann’s Bundle. Top, A postmortem human heart with a superior view at Bachmann’s bundle. Bottom, Color-coded activation maps illustrating the 4 different patterns of activation observed at Bachmann’s bundle. Propagation started either at the right (upper left map; N=124), right and central (upper right map; N=39), right, central, and left (lower left map; N=14), or solely from the central part (lower right map; N=8). Isochrones are drawn at 5 ms intervals, and arrows indicate the main direction of propagation. See text for detailed explanation. BB indicates Bachmann’s Bundle; I, inferior; L, left; LAA, left atrial appendage; LSPV, left superior pulmonary vein; R, right; S, superior; and SVC, superior caval vein.
Peak-to-Peak Amplitude of Sinus Rhythm Potentials
Mean voltages of all unipolar SR potentials (N=218±29/patient) were 3.0±1.4 mV and ranged from 0.3 to 7.2 mV. Lower averaged voltages were associated with ageing (P<0.001) and female sex (P=0.046); there were no correlations with a history of PAF, increased body mass index, hypertension, diabetes mellitus, hyperlipidemia, peripheral vascular disease, left ventricular dysfunction, or LA dilatation (P>0.05). Intraindividual variation in voltages across BB was 7.4±3.2 mV (minimum, 0.5±0.3 mV; maximum, 8.0±3.3 mV).
Conduction Velocity
The frequency distribution of the different effective conduction velocities of wavefronts propagating from the right to the left side of BB is shown in Figure 3 for the right side (left panel), central part (middle panel), and left side (right panel). In all patients, the effective conduction velocity did not differ between the right side (90±24 cm/s), central part (88±16 cm/s), or left side (89±15 cm/s) of BB (P>0.05); mean effective conduction velocity over the entire length of BB was 89±13 cm/s (range 57–128 cm/s). Lower conduction velocity was associated with lower voltages (P=0.002). Mean effective conduction velocity was not dependent on age (P=0.35) and was comparable between patients with PAF (97±15 cm/s) and patients without a history of AF (89±13 cm/s; P=0.09). Areas of slow conduction were observed in the majority of the patients (N=172; 93%); the median amount of slow conduction in all patients was 1.8% (0–9.2) and showed a trend toward a higher amount of slow conduction in patients with PAF compared with patients without a history of AF (2.6% [0.9–6.5] versus 1.7% [0–9.2]; P=0.07). Furthermore, a high amount of slow conduction (>2% or >4% slow conduction) was not associated with a higher age (P=0.16 and P=0.33).
Figure 3. Effective conduction velocity across Bachmann’s bundle. Relative frequency histograms in patients (N=124) with wave fronts propagating from right to left demonstrating the effective conduction velocity across Bachmann’s bundle in the right side (left), central part (middle), and left side (right).
Characteristics of Conduction Block
A frequency histogram of the amount of conduction block at BB is illustrated in the upper panel of Figure 4. In all patients, the median prevalence of conduction block was 1.1% and the mean prevalence was 1.9%. Areas of conduction block were present in the majority of the patients (N=138, 75%); in the remaining 47 patients (25%), conduction block did not occur. In patients with conduction block, the amount of conduction block varied from 0.2% to 12.8% (median prevalence: 1.8%). Conduction block was higher in patients with PAF compared with patients without a history of AF (3.2% [range 0–11.6] versus 0.9% [range 0–12.8]; P=0.03). Representative examples of the spatial distribution of areas of conduction block in patients with a variable amount of conduction block are depicted in the color-coded activation maps and corresponding conduction block maps in the middle panel of Figure 4. As can be seen in these maps, lines of conduction block occurred not only in the longitudinal direction, but also in the transverse direction of propagation. Electrograms around lines of conduction block showed both double potentials and fractionated potentials (Figure 4, lower panel).
Figure 4. Conduction block. Top, Frequency histogram demonstrating the incidence of conduction block across Bachmann’s bundle. Bottom, Color-coded activation maps and conduction block maps with a varying amount of conduction block ranging from 0.3% to 4.7%. Examples of epicardial unipolar potentials recorded from areas of conduction block are shown outside the map. The color-coded circles in front of the electrogram correspond to the areas of block indicated in the conduction block maps.
In the entire study population, the prevalence of longitudinal and transverse lines of conduction block ranged from, respectively, 0% to 12.8% (median: 1.3%) and 0% to 12.8% (median: 1.0%; P<0.01). Patients with PAF had a higher amount of conduction block in both longitudinal (1.1% [0–12.8] versus 4.0% [0–11.7]; P=0.03) and transverse direction (1.0% [0–12.8] versus 1.9% [0–12.3]; P=0.03).
As lines of longitudinal conduction block affect the right to left propagation across BB, the maximum lengths of all lines of longitudinal conduction block across BB were measured. Figure 5 shows the relative frequency of the maximal lengths of all longitudinal lines of conduction block observed in patients without a history of AF (left panel) or with PAF (right panel). Patients with PAF had longer lines of longitudinal conduction block than patients without a history of AF (median 8 mm versus 2 mm; P=0.03).
Figure 5. Longitudinal conduction block in Bachmann’s Bundle. Two pies illustrating relative incidence of the maximum lengths of lines of conduction block across Bachmann’s bundle in patients without a history of atrial fibrillation (AF; left) and with paroxysmal AF (PAF; right).
In most patients without a history of AF (51%), there were no or only small areas (2 mm) of longitudinal conduction block. Long lines of longitudinal conduction block (≥12 mm) were measured in only 12% of the patients (N=20). Although there were only 13 patients with PAF, solely 3 patients had no or small areas of conduction block. Five patients in this group (38%) had long lines of conduction block (≥12 mm). The maximum length of transverse lines of conduction block was also longer in patients with PAF than without AF (median 6 mm [range 0–20 mm] versus 2 mm [range 0–20 mm], P=0.03).
Impact of Longitudinal Conduction Block on Right to Left Propagation
The effects of longitudinal lines of conduction block on total activation time of BB and thus arrival time in the LA were determined for all patients mapped with the 192 unipolar mapping array with a single wavefront propagating from the right to left side of BB (N=52). The initial arrival times of these wavefronts at the LAA in relation to initial activation of BB for every patient individually are plotted in the upper panel of Figure 6. As can be seen, there is no effect of the length of the lines of longitudinal conduction block on the time required for right to left activation of BB.
Figure 6. Impact of longitudinal lines of conduction block. Top, The effect of the maximum length of line of conduction block in longitudinal direction and the first activation of the left atrial appendage is plotted for all patients (N=124) with a single right to left wavefront across Bachmann’s bundle. Bottom, Examples of the effect on the left atrial appendage activation time by different maximum lengths of lines of longitudinal conduction block. The left map illustrates a broad wavefront curving around the line of conduction block. The middle and right map demonstrate a right to left wavefront with a complete line of longitudinal conduction block coexisting with wavefronts entering BB from the central part (white or dashed asterisk) and left side.
Explanations for this observation are given in the lower panels of Figure 6. The left activation map shows a line of conduction block with a length of 12 mm with no effect on the right to left conduction. The wavefront propagated around the line of conduction block without any conduction delay and arrived at the LAA side 44ms after the first activation of BB. The middle and right activation maps demonstrate that even a complete line of conduction block (16 mm) across BB did not affect the arrival time at the LAA because in these patients areas behind the lines of conduction block were activated by wavefront emerging from other sites, including the left and central part of BB. As a result, the activation of the LAA site in these patients occurred only 33 ms and 22 ms after the first moment of activation of BB.
Early Postoperative Atrial Fibrillation
During the first 5 postoperative days, AF was observed in 56 patients (30%), including 7 patients (13%) who already had preoperative PAF. The incidence of de novo PoAF is plotted for each patient individually without a history of AF in Figure 7 and ranked according to the intraoperatively determined prevalence of longitudinal conduction block (x axis). There was a large variation in the length of areas of longitudinal conduction block in patients who developed PoAF. Although 10 patients (50%) with long lines of longitudinal conduction block developed PoAF, patients without or only small areas (2 mm) of conduction block in longitudinal direction also frequently developed PoAF (N=20; 23%).
Figure 7. Relation between the amount of longitudinal conduction block and development of early postoperative AF. Top, The effect of the amount of longitudinal conduction block on development of early postoperative AF in patients without a history of atrial fibrillation (AF; N=172). Bottom, Results of univariate and multivariate logistic regression between clinical determinants and development of early postoperative AF. CB indicates conduction block; CI, confidence interval; LA, left atrium; and OR, odds ratio.
The lower panel in Figure 7 shows the results of univariate and multivariate analyses. A higher age was not related with the occurrence of PoAF in patients without a history of AF (odds ratio 1.0, 95% confidence interval 1.0–1.1; P=0.31). Although an equal amount of conduction block was found in patients with PoAF compared with patients without PoAF (P=0.09, not shown in Figure 7), >4% conduction block was associated with development of PoAF (odds ratio 3.1, 95% confidence interval 1.2–8.1; P=0.02). When analyzing the amount of conduction block for the different orientations separately, there was no difference between the amount of transverse (P=0.06) or longitudinal conduction block (P=0.14) and development of PoAF. Also, a higher risk of development of PoAF was not associated with >4% conduction block in either longitudinal (P=0.28) or transverse direction (P=0.26).
In patients without a history of AF, the length of lines of longitudinal conduction block did not differ between patients with and without PoAF (median 4 mm versus 2 mm; P=0.07). However, patients with PoAF had more often long lines (≥12 mm) of longitudinal conduction block (N=11) compared with patients without PoAF (N=9; P<0.01). Patients with long lines of longitudinal conduction block had a 3 times higher risk (odds ratio 2.9; 95% confidence interval 1.1–8.2; P=0.04) of developing PoAF, whereas patients with lines of conduction block of 12 mm or longer in transverse direction had the same risk of developing PoAF (odds ratio 2.2; 95% confidence interval 0.51–9.9; P=0.28).
Discussion
High-resolution epicardial mapping of BB during SR in patients with coronary artery disease showed that BB was activated by multiple wavefronts entering BB not only from the right side, but also from the left side and central part in a considerable number of patients. The average effective conduction velocity across BB was ≈89 cm/s and did not differ between patients with or without AF. Lines of conduction block were found in the majority of the patients (74%) and occurred both in longitudinal and transverse direction. The effect of these lines of conduction block on excitation of the LA was limited. However, a high amount of conduction block and long lines of longitudinal conduction block were associated with the presence of PoAF.
Preferential, but Not the Only Interatrial Route
Experimental studies demonstrated that crushing of BB led to significant delay in excitation of the LA.1 However, in our study population, the presence of long lines of longitudinal conduction block did not result in delayed LAA activation because areas behind the lines of conduction block were activated by wavefronts emerging from either the left side or central part of BB.
As demonstrated in previous studies, our observations confirm that BB is not the exclusive route of interatrial conduction and that propagation of electric waves from the right to the LA occurs along other interatrial pathways when conduction across BB is impaired. These other interatrial pathways include the limbus of the fossa ovalis, the coronary sinus, and interatrial bundles both superior and inferior along BB.21–24
Conduction across BB has to date only indirectly been examined by using endocardial and epicardial mapping techniques.25–27 In patients who underwent catheter ablation of AF, 3-dimensional electro-anatomic (non)contact mapping techniques were used to examine the first LA activation site during SR. The earliest LA activation was frequently observed at the antero-superior LA, which was assumed to be the end of BB. Activation at this site was either solitary or simultaneously with other interatrial sites, which often included the postero-septal wall or the limbus of the fossa ovalis.26,27 Similar to these findings, we also observed that in the majority of our patients, a single wavefront propagated across BB from the right to the LA, which may initially activate the LA.
In 30% of our patients, BB was activated by wavefronts emerging in the central part of the mapping area. Although some studies observed that BB is isolated from the interatrial septum, others suggested that muscular connections between BB and the interatrial septum are present.8,24,28 It is therefore likely that when right to left conduction along BB is delayed, BB can also be excited by wavefronts conducting faster in other interatrial pathways (eg, limbus of fossa ovalis or coronary sinus), propagating upwards in the interatrial septum and activating the central area of BB.
Interestingly, wavefronts not only entered BB on the right side and propagated leftwards, but also entered on the left side and propagated rightwards. These left-sided wavefronts emerged both early and late in relation to the onset of activation of the right side of BB. This can be explained by the presence of the aforementioned additional bundles parallel to BB, crossing the roof of the LA.24 When wavefronts propagate faster across these parallel bands than BB, they can enter BB relatively early on the left side and collide with the right to leftwards propagating wavefronts. When conduction across BB is delayed, interatrial conduction occurs in other interatrial pathways, resulting in late excitation at the left side of BB, depending on the length and degree of conduction delay of the pathway taken. Besides that, a wavefront emerging from the left side in BB in the presence of a long line of conduction block could also be explained by turning of a wavefront around the end of the line of conduction block outside the mapping array. However, as the mapping array covers the entire width of BB, this is unlikely.
Bachmann’s Bundle, the Superconductor?
Propagation of wavefronts occurs faster in longitudinal than transverse direction. It is therefore assumed that longitudinal parallel orientation of the fibers in BB results in higher conduction velocity, thereby making BB a preferential route of interatrial conduction. In addition, some studies suggest that the fibers of BB have specific characteristics similar to components of the specialized Purkinje fibers, such as a higher resting membrane potential, rapid velocity upstroke of the action potential, distinct overshoot, and a broader phase 2 plateau.5,6 However, in contrast to the Purkinje cells, action potentials in BB abbreviated after application of acetylcholine, which suggests BB cardiomyocytes differ from the Purkinje cells of the cardiac conduction system.5 Altogether, these studies showed that cardiomyocytes of BB have cell characteristics similar to both the specialized conduction system and atrial cardiomyocytes.
The effective conduction velocity across BB measured in animal studies was often faster at BB than other atrial sites.4,5,9 Goodman et al performed mapping in a Langendorff-perfused canine heart with a 5-point electrode array.9 They observed a maximum conduction velocity of 300 cm/s in BB, which is comparable with conduction velocity in specialized Purkinje fibers. However, in our study population, we measured an effective conduction velocity of only 89 cm/s in BB, which is comparable with conduction velocities at other sites in the atria.29 Previous studies found higher effective conduction velocities across BB by measuring velocity between only a few points. Conduction velocity in BB could have been overestimated because wavefronts propagating from the right to the LA might fuse with wave fronts entering the central part of BB. This results in a large simultaneously activated areas, which could mimic fast propagation of a single wavefront between the first and last activated site. In our study population, the effective conduction velocity might also have been overestimated because of late merging of wave fronts arising from deeper layers. Only single wavefronts propagating from the right to left site of BB were chosen to minimize the risk of overestimation. Yet, despite the presence of only one single wavefront, a different angle of the wavefront entering BB, which is highly anisotropic in nature, can influence conduction velocity. On the contrary, areas of simultaneous activation (>200 cm/s) were interpreted as central entry sites of wavefronts propagating partially through deeper layers. They were sometimes observed after a line of conduction block, whereas they also collided with a right to left propagating wavefront without being separated by lines of conduction block. In the latter case, the conduction velocity could have been overestimated.
The Role of BB in the Pathophysiology of AF
Waldo et al made surgical lesions in BB of dogs and observed significant changes in the P wave morphology and duration.10 Delay in BB led to partial interatrial conduction block, whereas complete block of BB caused advanced interatrial conduction block which was characterized by biphasic P waves, particularly in the inferior leads on the surface ECG. Clinical studies have shown that advanced interatrial conduction block increases the risk of developing atrial tachyarrhythmias, including AF.12,13
The role of BB in initiation and perpetuation of AF has been investigated in animal studies.14,30,31 In the goat model of Allessie, initiation of AF episodes was preceded by atrial extrasystolic beats that were blocked at the middle of BB.14 Subsequently, re-excitation at the same side of the line of conduction block suggested re-entry in BB. Mapping during AF of both atria and the interatrial septum in a sterile pericarditis canine model revealed multiple unstable re-entry circuits involving the interatrial septum.30,31 As BB was the most commonly used interatrial pathway for these reentry circuits, the investigators suggested that BB is essential for perpetuation of AF.30 In this same canine model, complete transection of BB with radiofrequency ablation resulted in termination and noninducibility of AF.31
We observed multiple entry sites of BB during SR in patients with and without lines of conduction block. A line of conduction block across the entire width of BB did not result in delayed LA activation and the specific P wave alterations associated with development of AF. These findings differ from the earlier observations of P wave alterations after surgical transection of BB. In case of complete surgical transection of BB, other muscular connections, that is, interatrial septal pathways to BB, may also be disrupted.
According to these earlier studies, BB may play an important role in development of AF, although the exact mechanism remains unclear. Although over 20% of the patients without or with only a small area of conduction block developed PoAF, a high amount of conduction block and long lines of longitudinal conduction block were associated with de novo PoAF. These results indicate that the length of lines of conduction block facilitate reentry and hence development of AF. Yet, patients without (long) lines of conduction block developed both PAF and PoAF as well, suggesting that not only areas of conduction block in BB are involved in development of AF. The amount of conduction block at BB may merely reflect electric disease which is also present elsewhere in the atria. Other atrial sites may contain more extensive areas of conduction block and thus play a larger role in the pathophysiology of AF.
Becker microscopically examined BB, terminal crest, and pulmonary vein areas in 20 postmortem mainly known with coronary artery disease; 10 patients had a history of PAF.32 In all patients, fibro-fatty tissue and fibrotic patches were found, which may cause conduction disorders as a result of disruption of cell-to-cell connections.29,30 These histological changes were more common in patients with PAF, which may explain the higher amount of conduction block in patients with PAF in our study population. As all examined areas—pulmonary veins, terminal crest, and BB—were more affected in patients with PAF, delayed intra-atrial conduction predisposing to development of AF is probably the result of extensive conduction block throughout the atria and interatrial connections rather than conduction block across BB only. In addition, conduction disorder may be further impaired by, for example, atrial extrasystolic beats, which in turn initiate AF episodes.
Study Limitations
Mapping of BB is solely performed at the epicardial surface and does not provide any information of wavefronts propagating partially in deeper layers or emerging from other atrial sites. Hence, only the effective conduction velocity can be assessed. Also, the definition of slow conduction and conduction block remains arbitrary, and slow conduction cannot definitely be excluded. In the individual patient, the exact proportions of BB are unknown, and the mapping array might not always have covered the entire BB. However, previous studies demonstrated that the size of BB equals approximately the size of the mapping array. Also, our study did not provide any information on conduction properties at other atrial sites. Because of the small group of patients with a history of AF, comparison between patients with and without a history of AF was limited. In line with that, a lack of power might explain the absence of a significant relation between slowing of conduction velocity, a higher age, and development of PoAF.
Conclusions
High-resolution mapping of BB in humans with coronary artery disease during SR demonstrated that BB may be the preferential route of interatrial conduction, but can be activated from other directions as well. As a consequence, conduction disorders exclusively in BB have limited impact on LA excitation. Despite the longitudinal orientation of BB fibers, BB is not the superconductor as previously suggested. Conduction is blocked in both longitudinal and transverse direction in the majority of the patients. Conduction disorders, particularly long lines of longitudinal conduction block, are more pronounced in patients with AF episodes.
Sources of Funding
Dr N.M.S. de Groot is supported by grants from the Erasmus Medical Center fellowship, Dutch Heart Foundation (2012T0046), LSH-Impulse grant (40-43100-98-008), CoolSingel Foundation (no.212), Bayer and Boehringer Ingelheim.
Disclosures
None.
Footnotes
References
- 1.
Bachmann G . The inter-auricular time interval.Am J Physiol. 1916; 41:309–320.Google Scholar - 2.
James TN . The connecting pathways between the sinus node and A-V node and between the right and the left atrium in the human heart.Am Heart J. 1963; 66:498–508.CrossrefMedlineGoogle Scholar - 3.
Lemery R, Guiraudon G, Veinot JP . Anatomic description of Bachmann’s bundle and its relation to the atrial septum.Am J Cardiol. 2003; 91:1482–1485, A8.CrossrefMedlineGoogle Scholar - 4.
Dolber PC, Spach MS . Structure of canine Bachmann’s bundle related to propagation of excitation.Am J Physiol. 1989; 257(5 pt 2):H1446–H1457.MedlineGoogle Scholar - 5.
Wagner ML, Lazzara R, Weiss RM, Hoffman BF . Specialized conducting fibers in the interatrial band.Circ Res. 1966; 18:502–518.LinkGoogle Scholar - 6.
Childers RW, Merideth J, Moe GK . Supernormality in Bachmann’s bundle. An in vitro and in vivo study in the dog.Circ Res. 1968; 22:363–370.CrossrefMedlineGoogle Scholar - 7.
Ho SY, Anderson RH, Sánchez-Quintana D . Atrial structure and fibres: morphologic bases of atrial conduction.Cardiovasc Res. 2002; 54:325–336.CrossrefMedlineGoogle Scholar - 8.
Platonov PG, Mitrofanova L, Ivanov V, Ho SY . Substrates for intra-atrial and interatrial conduction in the atrial septum: anatomical study on 84 human hearts.Heart Rhythm. 2008; 5:1189–1195. doi: 10.1016/j.hrthm.2008.04.025.CrossrefMedlineGoogle Scholar - 9.
Goodman D, van der Steen AB, van Dam RT . Endocardial and epicardial activation pathways of the canine right atrium.Am J Physiol. 1971; 220:1–11.MedlineGoogle Scholar - 10.
Waldo AL, Bush HL, Gelband H, Zorn GL, Vitikainen KJ, Hoffman BF . Effects on the canine P wave of discrete lesions in the specialized atrial tracts.Circ Res. 1971; 29:452–467.LinkGoogle Scholar - 11.
Agarwal YK, Aronow WS, Levy JA, Spodick DH . Association of interatrial block with development of atrial fibrillation.Am J Cardiol. 2003; 91:882.CrossrefMedlineGoogle Scholar - 12.
Ariyarajah V, Fernandes J, Kranis M, Apiyasawat S, Mercado K, Spodick DH . Prospective evaluation of atrial tachyarrhythmias in patients with interatrial block.Int J Cardiol. 2007; 118:332–337. doi: 10.1016/j.ijcard.2006.07.021.CrossrefMedlineGoogle Scholar - 13.
Bayes de Luna A, Cladellas M, Oter R, Torner P, Guindo J, Marti V, Rivera I, Iturralde P . Interatrial conduction block and retrograde activation of the left atrium and paroxysmal supraventricular tachyarrhythmia.Eur Heart J. 1988; 9:1112–1118.CrossrefMedlineGoogle Scholar - 14.
Duytschaever M, Danse P, Eysbouts S, Allessie M . Is there an optimal pacing site to prevent atrial fibrillation?: an experimental study in the chronically instrumented goat.J Cardiovasc Electrophysiol. 2002; 13:1264–1271.CrossrefMedlineGoogle Scholar - 15.
van Campenhout MJ, Yaksh A, Kik C, de Jaegere PP, Ho SY, Allessie MA, de Groot NM . Bachmann’s bundle: a key player in the development of atrial fibrillation?Circ Arrhythm Electrophysiol. 2013; 6:1041–1046. doi: 10.1161/CIRCEP.113.000758.LinkGoogle Scholar - 16.
Yaksh A, Kik C, Knops P, Roos-Hesselink JW, Bogers AJ, Zijlstra F, Allessie M, de Groot NM . Atrial fibrillation: to map or not to map?Neth Heart J. 2014; 22:259–266. doi: 10.1007/s12471-013-0481-0.MedlineGoogle Scholar - 17.
de Groot NM, Houben RP, Smeets JL, Boersma E, Schotten U, Schalij MJ, Crijns H, Allessie MA . Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough.Circulation. 2010; 122:1674–1682. doi: 10.1161/CIRCULATIONAHA.109.910901.LinkGoogle Scholar - 18.
Allessie MA, de Groot NM, Houben RP, Schotten U, Boersma E, Smeets JL, Crijns HJ . Electropathological substrate of long-standing persistent atrial fibrillation in patients with structural heart disease: longitudinal dissociation.Circ Arrhythm Electrophysiol. 2010; 3:606–615. doi: 10.1161/CIRCEP.109.910125.LinkGoogle Scholar - 19.
Rogers JM, Usui M, KenKnight BH, Ideker RE, Smith WM . Recurrent wavefront morphologies: a method for quantifying the complexity of epicardial activation patterns.Ann Biomed Eng. 1997; 25:761–768.CrossrefMedlineGoogle Scholar - 20.
Kay MW, Gray RA . Measuring curvature and velocity vector fields for waves of cardiac excitation in 2-D media.IEEE Trans Biomed Eng. 2005; 52:50–63. doi: 10.1109/TBME.2004.839798.CrossrefMedlineGoogle Scholar - 21.
Sanchez-Quintana D, Davies DW, Ho SY, Oslizlok P, Anderson RH . Architecture of the atrial musculature in and around the triangle of Koch: its potential relevance to atrioventricular nodal reentry.J Cardiovasc Electrophysiol. 1997; 8:1396–1407.CrossrefMedlineGoogle Scholar - 22.
Mitrofanova L, Ivanov V, Platonov PG . Anatomy of the inferior interatrial route in humans.Europace. 2005; 7(suppl 2):49–55. doi: 10.1016/j.eupc.2005.03.014.CrossrefMedlineGoogle Scholar - 23.
Chauvin M, Shah DC, Haïssaguerre M, Marcellin L, Brechenmacher C . The anatomic basis of connections between the coronary sinus musculature and the left atrium in humans.Circulation. 2000; 101:647–652.CrossrefMedlineGoogle Scholar - 24.
Ho SY, Sanchez-Quintana D, Cabrera JA, Anderson RH . Anatomy of the left atrium: implications for radiofrequency ablation of atrial fibrillation.J Cardiovasc Electrophysiol. 1999; 10:1525–1533.MedlineGoogle Scholar - 25.
Markides V, Schilling RJ, Ho SY, Chow AW, Davies DW, Peters NS . Characterization of left atrial activation in the intact human heart.Circulation. 2003; 107:733–739.LinkGoogle Scholar - 26.
Tapanainen JM, Jurkko R, Holmqvist F, Husser D, Kongstad O, Mäkijärvi M, Toivonen L, Platonov PG . Interatrial right-to-left conduction in patients with paroxysmal atrial fibrillation.J Interv Card Electrophysiol. 2009; 25:117–122. doi: 10.1007/s10840-008-9359-2.CrossrefMedlineGoogle Scholar - 27.
Lemery R, Birnie D, Tang AS, Green M, Gollob M, Hendry M, Lau E . Normal atrial activation and voltage during sinus rhythm in the human heart: an endocardial and epicardial mapping study in patients with a history of atrial fibrillation.J Cardiovasc Electrophysiol. 2007; 18:402–408. doi: 10.1111/j.1540-8167.2007.00762.x.CrossrefMedlineGoogle Scholar - 28.
Papez JW . Heart musculature of the atria.Am J Anat. 1920–21; 27:255–277.CrossrefGoogle Scholar - 29.
Hansson A, Holm M, Blomström P, Johansson R, Lührs C, Brandt J, Olsson SB . Right atrial free wall conduction velocity and degree of anisotropy in patients with stable sinus rhythm studied during open heart surgery.Eur Heart J. 1998; 19:293–300.CrossrefMedlineGoogle Scholar - 30.
Kumagai K, Khrestian C, Waldo AL . Simultaneous multisite mapping studies during induced atrial fibrillation in the sterile pericarditis model. Insights into the mechanism of its maintenance.Circulation. 1997; 95:511–521.CrossrefMedlineGoogle Scholar - 31.
Kumagai K, Uno K, Khrestian C, Waldo AL . Single site radiofrequency catheter ablation of atrial fibrillation: studies guided by simultaneous multisite mapping in the canine sterile pericarditis model.J Am Coll Cardiol. 2000; 36:917–923.CrossrefMedlineGoogle Scholar - 32.
Becker AE . How structurally normal are human atria in patients with atrial fibrillation?Heart Rhythm. 2004; 1:627–631. doi: 10.1016/j.hrthm.2004.09.009.CrossrefMedlineGoogle Scholar
