Human Study of Biatrial Electrical Coupling
Determinants of Endocardial Septal Activation and Conduction Over Interatrial Connections
Background— The relative contribution of the atrial septum and interatrial connections to biatrial activation is a fundamental concept of human cardiac electrophysiology that has yet to be fully characterized. The purpose of the present study was to determine how both atria are coupled electrically.
Methods and Results— Twenty patients (16 men; mean age 54±11years) with a history of symptomatic atrial fibrillation (AF) underwent simultaneous biatrial noncontact mapping before catheter ablation of AF. The multiple electrode array catheters were positioned, respectively, in the left atrium (LA; transseptally) and the right atrium (RA). In all but 2 patients, isopotential maps revealed that endocardial septal activations of the RA and LA were separate, independent, and asynchronous of each other. Interatrial conduction was related to the site of initial atrial depolarization, revealing conduction over Bachmann’s bundle in all patients during sinus rhythm, high RA pacing, and pacing from the LA appendage. Pacing from the coronary sinus was associated with conduction over the interatrial connection at the level of the coronary sinus in all patients, and conduction over Bachmann’s bundle also occurred in 5 (26%) of 19 patients. Interatrial conduction over the fossa ovalis occurred in only 2 (2%) of the 116 segments analyzed.
Conclusions— Electrical coupling of the RA and LA in humans is predominantly provided by muscular connections at the level of Bachmann’s bundle and the coronary sinus. The true septum (the fossa ovalis and its limbus) of the RA and LA is asynchronous and discordant, usually without contralateral conduction during sinus rhythm or atrial pacing.
Received January 21, 2004; revision received March 16, 2004; accepted March 19, 2004.
The anatomic characteristics of interatrial connections in humans have been described recently.1–4 Both Bachmann’s bundle4 and the coronary sinus2 provide muscular connections between both atria; however, their distinct conduction properties and their interaction have yet to be specifically characterized in humans.5 The true interatrial septum (the fossa ovalis and its limbus) is generally viewed as a common anatomic separation of both atria, or with the general notion that electrophysiological continuity at the level of the fossa ovalis is usually present.6
Nonfluoroscopic mapping systems provide 3D reconstruction of heart chambers and allow for real-time continuous acquisition of electrophysiological activation.6–8 In particular, noncontact mapping detects far-field endocardial potentials of a cardiac chamber, providing unipolar (virtual) electrograms at 3360 points, which produces dynamic isopotential maps. Thus, the spread of isopotential wavefronts and precise determination of interatrial breakthrough sites can be obtained.7,8
Simultaneous biatrial noncontact mapping in humans has not been reported previously. We performed such a study in patients and provide for the first time in humans a detailed analysis of biatrial endocardial septal activation, as well as electrophysiological characteristics of conduction over interatrial connections in sinus rhythm and during atrial pacing.
Twenty patients (16 men; mean age 54±11 years) underwent simultaneous biatrial noncontact mapping at the time of an electrophysiological study and before catheter ablation for symptomatic AF. The Institutional Review Board of the University of Ottawa Heart Institute approved the study. All patients gave written informed consent and were studied in the fasting postabsorptive state. Antiarrhythmic medications were stopped >5 half-lives before the electrophysiological study, except for 8 patients receiving amiodarone, in whom the drug was stopped 1 week before their procedure. All patients were receiving warfarin, which was stopped 5 days before the study. A transesophageal echocardiogram was done within 2 days of the study; an intracardiac thrombus was excluded in all patients. Table 1 outlines the characteristics of the patients.
Noncontact Mapping System
The noncontact mapping system (Endocardial Solutions Inc) was used to create the right atrial (RA) and left atrial (LA) activation maps. The system has been described previously.7,8 The multiple electrode array (MEA) catheter consists of a 7.5-mL balloon mounted on a 9F catheter. The system displays a color dynamic 3D isopotential map derived from the amplitude of the unipolar (virtual) electrogram at 0.83-ms intervals. With manual adjustments and predominant filtering of unipolar electrograms at 2 Hz, the leading-edge voltage of the depolarization wavefront (negative potential) could be observed and followed.
Positioning of Catheters and Creation of Atrial Geometries
In all patients, catheters were advanced under fluoroscopic guidance through sheaths introduced in the right and left femoral veins under local anesthesia (Figure 1). A double-transseptal approach was performed with an SL-1 Daig (St. Jude Medical) sheath and the Brockenbrough needle for each transseptal puncture. Heparin was administered to maintain activated clotting times of ≥300 seconds throughout the study. The distal pigtail tip of the MEA was usually positioned in the appendage region for the LA and in the ostial region of the superior vena cava for the RA (both of which remained stable during the mapping study). Specific anatomic locations in both atrial chambers during pacing or mapping were confirmed by orientation of the fluoroscopic image in the anteroposterior view and at 30° to 40° of both the right anterior oblique and left anterior oblique views. Bipolar recordings were recorded between a band pass of 30 to 500 Hz and used to confirm local capture and biatrial propagation patterns. Surface ECG leads I, II, and V1 were recorded continually throughout the study.
All patients underwent simultaneous deployment of both MEAs (2 hours 32 minutes±1 hour 14 minutes) without electrical or hemodynamic instability. However, there were femoral complications, including a small AV fistula in 2 patients (both were asymptomatic), and 2 patients developed a femoral hematoma that resolved.
Both MEAs were connected into their respective breakout boxes, and bipolar and unipolar electrograms and isopotential maps were projected on their respective graphic workstations. In all patients, great care was taken to ensure that the RA and LA septum could be aligned properly. First, the fossa ovalis could be delineated carefully because of the reference transseptal puncture site and the shaft of the MEA going through the fossa ovalis to the LA. In the first 4 patients, the fossa ovalis was used to circumscribe the center of the region overlying the true common septum. For all remaining patients, in addition to using the fossa ovalis area as a reference, the RA and LA catheters were respectively positioned to 3 separate points around the region of the fossa. The Enguide feature (ie, the catheter location signal of the Ensite) of each mapping system was steered to these respective septal points. A Docking software was designed that used paired septal points and triangular alignment to create the septal region of the RA and LA in contact with each other. Thus, the atria could be aligned side by side, with the fossa ovalis centering this region.9 During subsequent data analysis sessions, both atrial chamber images were aligned, animated, and analyzed with Macromedia Flash software.
Simultaneous Analysis of Both Atria
Precise and simultaneous analysis of electrograms of each noncontact mapping system was ensured by using a digital marker signal sent to both workstations that provided for a maximum delay of 0.83 ms.
Sinus Rhythm Analysis and Atrial Pacing Protocol
Activation maps were obtained in sinus rhythm and during stable and regular atrial pacing sites, the latter confirmed by multiple fluoroscopic views. Pacing was performed at twice-diastolic thresholds, usually at cycle lengths between 500 and 700 ms, or exceeding the underlying sinus rate by 100 ms. Activation time was defined as the duration in milliseconds after earliest endocardial breakout from the region being evaluated (after sinus node depolarization or endocardial capture during pacing) to the earliest endocardial activation of the region of interest being analyzed. The LA and RA insertions of Bachmann’s bundle were determined, respectively, by earliest LA endocardial activation in the region of Bachmann’s bundle during sinus rhythm and by earliest RA activation in the region of Bachmann’s bundle during pacing from the roof of the LA near the predetermined region of Bachmann’s bundle in sinus rhythm.
Values are reported as mean±SD. Unpaired 2-tailed t test for variance was used to determine statistical significance. A probability value of <0.01 was considered statistically significant.
In sinus rhythm (Table 2), P-wave duration varied between 71 and 141 ms (mean 101±16 ms). The RA insertion of Bachmann’s bundle was activated significantly earlier than the LA insertion of Bachmann’s bundle (19±12 versus 41±18 ms, P<0.00004) and occurred before RA septal activation in 14 (70%) of 20 patients. In all patients, LA activation occurred over Bachmann’s bundle (mean conduction time over Bachmann’s bundle was 23±15 ms).
Mean septal activation time was significantly longer in the RA than the LA (48±14 versus mean 33±9 ms; P<0.001). In all but 2 patients, isopotential maps revealed that endocardial septal activation of the RA and LA was separate, independent, and asynchronous of each other. As shown in Figure 2, unipolar electrograms revealed propagation of endocardial activation of the right septum that preceded the left septum, and activation of one side did not influence the other. As the wavefront moved across the fossa ovalis, neither side activated the contralateral side.
The sequence of activation of septal activation and type of interatrial connection for pacing of the high RA closely resembled observations seen in sinus rhythm. When pacing the low lateral RA, endocardial conduction time to septal activation was greatly delayed (mean values in Table 2). LA activation occurred only over Bachmann’s bundle in 15 (79%) of 19 patients, only over the coronary sinus in 2 patients, over both the coronary sinus and Bachmann’s bundle in 1 patient, and over the fossa ovalis only in 1 patient.
When pacing the LA appendage, interatrial conduction and observations on septal activation were the opposite of those observed during high RA pacing. When pacing the coronary sinus, either proximal or distal, interatrial conduction occurred over the coronary sinus region in all but 1 patient, who had conduction over Bachmann’s bundle only when distal pacing from the coronary sinus was performed. In addition, a second interatrial connection site over Bachmann’s bundle was documented in 5 (26%) of 19 patients when proximal CS pacing was performed and in 3 (16%) of 19 when distal CS pacing was performed. During the latter, there was also 1 patient who had conduction over both the fossa ovalis and the coronary sinus. Distal coronary sinus pacing (Figure 3) was associated with significantly longer endocardial activation times to reach the LA septum than proximal coronary sinus pacing (71±21 versus 28±30 ms, P<0.0009).
Effect of Amiodarone
There were 8 patients in whom amiodarone was stopped 1 week before the electrophysiological study. The longest conduction times documented both during sinus rhythm and during atrial pacing were recorded in patients 8 and 9. Mean conduction times of the other patients receiving amiodarone were not significantly different from all other patients.
The present results demonstrate that whether the septum was initially activated from the RA or LA or from the superior or inferior aspect, separate and independent wavefront activations were observed, usually without interatrial conduction over the true septum. In sinus rhythm, conduction over Bachmann’s bundle (mean 23±15 ms) provides the requirements for nearly synchronous biatrial activation.
Biatrial mapping of atrial arrhythmias, performed in both animals and humans, have used epicardial and endocardial techniques to study septal activation.10–17 These reports suggested delayed LA septal activation compared with RA activation in sinus rhythm,12,14,17 as well as asynchronous septal activation in the dog.5 A detailed review of biatrial mapping was recently reported.18
Recent anatomic studies have revealed the dimensions of muscular connections at the level of the coronary sinus2 and Bachmann’s bundle.4 The present report clearly demonstrates the importance of these pathways during interatrial conduction. The septum primum at the level of the fossa ovalis has been shown to average only 1.8±0.7 mm in thickness, whereas the anterior and posterior regions of the septum average 6±2 mm and range from 2 to 20 mm in thickness.19 Histological cross sections of the midseptal wall have shown connective tissue, fat, and vessels separating the RA and LA septal surfaces.5
Partial Insulation of Both Atria
As reviewed by Schuessler et al,5 a detailed understanding of human atrial electrical coupling was not available previously. The present study helps define and clarify 2 important features of endocardial atrial activation. First, whether the septum is activated in the caudocranial or craniocaudal direction, interatrial conduction over the true septum generally does not occur, likely owing to an anisotropy and septal fiber orientation. Second, although interatrial conduction is provided by muscular connections at the level of the coronary sinus and Bachmann’s bundle,15–17,20,21 the electrophysiological characteristics of the latter are unique. Indeed, whether Bachmann’s bundle was activated during sinus rhythm or atrial pacing, mean conduction times varied little, between 17 and 23 ms. As shown by Bachmann in dogs,22 the role of this interatrial band is to ensure rapid interatrial conduction, leading to physiological biatrial contraction (thus superseding any need for interatrial conduction over the true septum).
Disease states23,24 or selective ablation of Bachmann’s bundle25,26 can cause an increase in interatrial or biatrial conduction times. Under these circumstances, interatrial conduction over the true septum or coronary sinus may occur, but significant biatrial delay may cause nonphysiological or nonsynchronous atrial contraction.23,27
The present report outlines biatrial septal activation and patterns of interatrial connections in patients with paroxysmal or persistent AF. These patients were predominantly healthy males (75%), with a mean height and weight of 182±6 cm and 91±15 kg, respectively. Mean 2D echocardiographic LA diameter was 43±6 mm. We do not know whether similar observations would be present in patients without arrhythmia disorders or in those with advanced heart disease. Other mapping and anatomic studies have shown other, less predominant types of interatrial conduction.1,28 We attempted to regroup interatrial conduction under 3 predominant regions but cannot exclude that other smaller interatrial connections may have been present in some patients. Finally, septal pacing was not performed, to prevent far-field capture of the contralateral septum or other atrial chamber. Also, transseptal access may have affected conduction over the true septum, but this would not be suggested by the findings of previous epicardial studies in dogs.5
Electrical coupling of both atria in humans demonstrates asynchronous and discordant atrial septal activation, both in sinus rhythm and during atrial pacing. In the majority of patients, only ipsilateral conduction over the true septum is present. Interatrial connections at the level of Bachmann’s bundle and the coronary sinus provide the requirements for biatrial activation during both normal physiological conditions and ectopic rhythms.
Chauvin M, Shah DC, Haissaguerre M, et al. The anatomic basis of connections between the coronary sinus musculature and the left atrium in humans. Circulation. 2000; 101: 647–652.
Anderson RH, Brown NA, Webb S. Development and structure of the atrial septum. Heart. 2002; 88: 104–110.
Schuessler RB, Boineau JP, Bromberg BI, et al. Normal and abnormal activation of the atrium. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology: From Cell to Bedside. 2nd ed. Philadelphia, Pa: WB Saunders; 1995: 543–562.
Roithinger FX, Cheng J, Groenewegen S, et al. Use of electroanatomic mapping to delineate transseptal atrial conduction in humans. Circulation. 1999; 100: 1791–1797.
Schilling RJ, Peters NS, Davies DW. Simultaneous endocardial mapping in the human left ventricle using a non-contact catheter: comparison of contact and reconstructed electrograms during sinus rhythm. Circulation. 1998; 98: 887–898.
Kadish A, Hauck J, Pederson B, et al. Mapping of atrial activation with a non-contact multielectrode catheter in dogs. Circulation. 1999; 99: 1906–1913.
Antz M, Otomo K, Arrunda M, et al. Electrical conduction between the right atrium and left atrium via the musculature of the coronary sinus. Circulation. 1998; 98: 1790–1795.
Marine JE, Korley VJ, Obioha-Ngwu O, et al. Different patterns of interatrial conduction in clockwise and counterclockwise atrial flutter. Circulation. 2001; 104: 1153–1157.
Sun H, Velipasaoglu EO, Wu DE, et al. Simultaneous multisite mapping of the right and the left atrial septum in the canine intact beating heart. Circulation. 1999; 100: 312–319.
Boineau JP, Canavan TE, Schuessler RB, et al. Demonstration of a widely distributed atrial pacemaker complex in the human heart. Circulation. 1988; 77: 1221–1237.
Rodriguez LM, Timmermans C, Nabar A, et al. Biatrial activation in isthmus-dependent atrial flutter. Circulation. 2001; 104: 2545–2550.
Bachmann JG. The inter-auricular time interval. Am J Physiol. 1916; 41: 309–320.
Waldo AL, Bush HL, Gelband H, et al. Effects on the canine P wave of discrete lesions in the specialized atrial tracts. Circ Res. 1971; 29: 452–461.
Markides V, Schilling RJ, Ho SY, et al. Characterization of left atrial activation in the intact human heart. Circulation. 2003; 107: 733–739.