Author + information
- Received June 18, 2007
- Revision received October 29, 2007
- Accepted November 8, 2007
- Published online February 26, 2008.
- Kurt C. Roberts-Thomson, MBBS⁎,1,
- Irene H. Stevenson, MBBS⁎,2,
- Peter M. Kistler, MBBS, PhD⁎,3,
- Haris M. Haqqani, MBBS⁎,1,
- John C. Goldblatt, MBBS⁎,
- Prashanthan Sanders, MBBS, PhD† and
- Jonathan M. Kalman, MBBS, PhD, FACC⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Jonathan M. Kalman, Department of Cardiology, Royal Melbourne Hospital, Royal Pde. Melbourne 3050, Australia.
Objectives This study sought to characterize the conduction properties of the posterior left atrium (PLA) in patients with different forms of structural heart disease undergoing cardiac surgery.
Background The PLA plays an important role in the initiation and maintenance of atrial fibrillation.
Methods This study included 34 patients having elective cardiac surgery. There were 4 groups of patients: normal left ventricular (LV) function (coronary artery bypass grafting [CABG]); severe LV dysfunction (LVF/CABG); severe mitral regurgitation (MR); severe aortic stenosis (AS). Epicardial mapping of the PLA was performed in sinus rhythm and during differential pacing. Activation patterns, regional conduction velocity (CV), conduction heterogeneity, anisotropy, and total plaque activation time (TAT) were assessed.
Results Left atrial size in patients with LVF/CABG (47 ± 7 mm) and MR (54 ± 6 mm) was larger than patients with CABG (39 ± 7 mm) and AS (42 ± 6 mm; p < 0.05). During pacing, all patients developed a vertical line of conduction delay running between the pulmonary veins. The extent of this conduction delay was greater in patients with LVF/CABG and MR than patients with AS and CABG (p < 0.05). Conduction heterogeneity, anisotropy, and TAT were greater in patients with LVF/CABG and MR than patients with CABG (p < 0.05). These changes resulted in circuitous wave front propagation.
Conclusions There is a line of functional conduction delay in a consistent anatomical location in the PLA in patients with structural heart disease. This is most marked in conditions associated with significant chronic atrial enlargement and leads to circuitous wave front propagation, suggesting a potential role in arrhythmogenesis.
Over the past decade, a spectrum of mapping, ablation, and surgical studies have suggested an important role for the posterior wall of the left atrium (PLA) both in the initiation and the maintenance of atrial fibrillation (AF) (1–6). Although early reports showed the importance of pulmonary vein musculature in the initiation of AF (1), more recent studies have emphasized the critical and dynamic interplay between the pulmonary veins and PLA musculature (4,6,7). The role of the PLA in maintaining AF was also highlighted by Todd et al. (3), who successfully treated AF by surgically isolating the PLA en bloc. Further, postoperative electrophysiological study showed spontaneous or inducible AF in the isolated PLA, whereas the rest of the left atrium (LA) was unable to sustain AF.
Recently, Markides et al. (8) used noncontact mapping to describe the electrophysiological properties of this region in patients with paroxysmal AF. They showed the presence of a line of functional conduction block, descending the PLA wall, between the superior and inferior pulmonary veins, which correlated with a region of change in subendocardial fiber orientation observed in pathology specimens. During the initiation of AF, this line of block formed the substrate for functional re-entry, causing wave fronts of pulmonary vein origin to divide into daughter wavelets.
The goal of the present study was to further characterize the electrophysiological properties of this PLA region in patients with different forms of structural heart disease. We hypothesized that structural heart disease would amplify conduction slowing and heterogeneity in this region and create the substrate for re-entry.
This study included 34 patients with stable coronary artery disease or valvular heart disease undergoing elective cardiac surgery at our institution. Patients were included in 1 of 4 groups:
Group A (coronary artery bypass grafting [CABG]): patients with normal left ventricular function undergoing CABG (8 patients)
Group B (left ventricular failure [LVF]/CABG): patients with severe left ventricular dysfunction (left ventricular ejection fraction <35%) undergoing CABG (10 patients)
Group C (mitral regurgitation [MR]): patients with normal left ventricular function undergoing mitral valve surgery for severe MR (8 patients)
Group D (aortic stenosis [AS]): patients with normal left ventricular function undergoing aortic valve replacement for severe AS (8 patients)
Patients in the above groups were in sinus rhythm with no clinical history or documented evidence of atrial arrhythmias.
All antiarrhythmic medications were ceased >5 half lives before surgery. All patients gave written informed consent before their surgery, and the study protocol was approved by the research and ethics committee of Melbourne Health.
Mapping and pacing protocol
After median sternotomy and before the institution of cardiopulmonary bypass, atrial epicardial mapping was performed. A custom-made high-density epicardial plaque (128 silver-plated copper electrodes with an interelectrode distance of 2.5 mm) was placed in the oblique sinus on the PLA between the pulmonary veins.
Isochronal maps were generated during sinus rhythm and during pacing from each of the 3 corners (1 superior and 2 inferior) of the epicardial plaque at cycle lengths of 600 and 400 ms.
Bipolar atrial electrograms were recorded and stored on a computerized mapping system (Unemap, Uniservices, Auckland, New Zealand) for offline analysis. Isochronal maps were then generated with 3-ms intervals in local activation time to assess conduction and activation patterns.
Regional conduction: conduction slowing and conduction block
Assessment of regional conduction slowing was performed by triangulating the plaque electrodes as previously described (9,10). Local conduction velocity was calculated as the mean of 2 conduction vectors in a 2.5 × 2.5-mm area. The percentage of the epicardial plaque with conduction slowing or conduction block was used to define the extent of regional conduction slowing or block. Regional conduction slowing was assessed during pacing from the superior and inferior plaque corners at each of the pacing cycle lengths.
Total plaque activation time and conduction anisotropy
The effect of conduction directionality on conduction time was assessed. Total activation time of the plaque was assessed for each of the pacing sites and cycle lengths. This provided an assessment of plaque activation times for wave fronts propagating in both vertical and horizontal directions. The anisotropy index was calculated by dividing the longest activation time by the shortest activation time (10).
Heterogeneity of conduction was determined using the phase mapping method described by Lammers et al. (11). In brief, the largest activation time difference between 4 adjacent electrodes is recorded. These activation time differences are then plotted on a histogram for the entire plaque. Absolute heterogeneity is calculated by subtracting the 5th percentile from the 95th percentile (P95 − P5) of the activation time differences. The conduction heterogeneity index is then determined by dividing this number by the median (P50) of the activation time differences.
Conduction heterogeneity was assessed during sinus rhythm and pacing from the superior and inferior plaque corners at each of the pacing cycle lengths.
Atrial effective refractory period (AERP)
The AERP was measured at the superior corner of the plaque at cycle lengths of 600 and 400 ms at twice diastolic threshold. An incremental technique was used, starting with an S2 coupling interval of 150 ms, increased in intervals of 10 ms. The AERP was defined as the longest coupling interval that failed to propagate to the atrium. The AERP was measured 2 times at each cycle length, and if the maximum and minimum measurements differed by >10 ms, then an extra measurement was performed and the total averaged.
Double potentials were defined as 2 potentials separated by an isoelectric interval. Fractionated electrograms were defined as electrograms of >50 ms duration with more than 3 deviations from baseline (12). Slow conduction was defined as local conduction velocities between 10 and 20 cm/s (9). Conduction block was defined as local conduction velocities of ≤10 cm/s (9).
Uniform conduction across the plaque was considered to be present when there was even distribution of isochrones across the plaque and there were no regions that met the criteria for slow conduction or conduction block.
All results were expressed as mean ± standard deviation. Comparisons between multiple groups, pacing sites, and pacing cycle lengths were performed using analysis of variance, with post hoc analysis using a Bonferroni correction for multiple comparisons. Proportions were compared using chi-square tests. Relationships between variables were evaluated using a Pearson correlation coefficient. A value of p < 0.05 was considered statistically significant.
The baseline characteristics of the patients are presented in Table 1.
Activation patterns: sinus rhythm
Earliest activation of the plaque during sinus rhythm occurred at the right inferior corner in 26 patients and at the superior aspect of the plaque in 8 patients. In 29 patients activation maps showed uniform conduction across the PLA during sinus rhythm. In 5 patients, a line of isochronal crowding meeting the criteria for a zone of slow conduction was present, running vertically down the PLA wall. This line of slow conduction was only observed when earliest activation occurred at the right inferior corner of the plaque with propagation in a transverse direction across the PLA wall. There were no significant differences observed between the groups for PLA activation in sinus rhythm.
Activation patterns: pacing from the plaque corners
During pacing from corners of the plaque, all patients developed a vertical line of slow conduction running in a highly stereotypic anatomical location (Figs. 1 and 2).⇓⇓ The line descended on the posterior wall of the LA from the roof, passing between the ostia of the superior and then between the ostia of the inferior pulmonary veins. Local double potentials or fractionated signals were always present in the region of slow conduction (Fig. 1). During pacing, there were 3 different types of plaque activation pattern observed according to the extent of conduction slowing or conduction block along this line: 1) slowed conduction with crowding of isochrones along the length of the line but no regions of conduction block (Fig. 3); 2) incomplete conduction block along the line with 1 or more regions where the wave front breaks across the line at a point of slowed conduction (Figs. 1B and 2C); and 3) complete conduction block along the length of the line with the wave front forced to take a circuitous path around the line (Figs. 1A, 1C, 2A, and 2B). In 7 patients, the line of slowed conduction was only present when pacing 2 of the 3 corners.
Regional conduction slowing or conduction block
Data showing the percentage of the plaque with conduction slowing or block during pacing from the plaque corners are shown in Table 2. The extent of conduction slowing and block was significantly greater in Group B (LVF/CABG) and Group C (MR) patients than in Group A (CABG) or Group D (AS) patients (p < 0.01). There was no significant difference between Group D (AS) and Group A (CABG) patients. Further, the percentage of conduction slowing and block was greater at a pacing cycle length of 400 than 600 ms (10.3 ± 7.2% vs. 8.5 ± 6.8%; p < 0.05). There was no difference when pacing the superior corner compared with the inferior corners of the plaque (9.5% ± 8.0% vs. 8.9% ± 7.0%; p = 0.34). The proportion of patients with conduction block was greater in Group B (LVF/CABG), Group C (MR), and Group D (AS) than in Group A (CABG) (Group A 12%, Group B 45%, Group C 37%, Group D 32%; p < 0.001). There was a significant correlation between LA diameter and the percentage of the plaque with conduction slowing or block during pacing (R = 0.36, p < 0.05).
Distribution of conduction slowing and conduction block
In all patients, the distribution of the region of slow conduction and conduction block was in a vertical line running between the pulmonary veins. Because of variations in pericardial anatomy, plaque position altered slightly between patients and so the region of slow conduction varied in position on the plaque. However, within an individual patient, this region was highly stereotypic and remained stable across different pacing sites, cycle lengths, and when observed during sinus rhythm (Figure 2).
Total atrial conduction time
Data for total atrial conduction time when pacing the inferior plaque corner are shown in Table 3. Activation time was significantly longer in Group B (LVF/CABG) and Group C (MR) than in Group A (CABG) or Group D (AS) at both pacing cycle lengths (p < 0.01).
The results are shown in Table 4. In all patients, conduction time in a vertical direction (parallel to the line of slow conduction) was shorter than conduction time in a horizontal direction (perpendicular to the line of slow conduction). Patients in Group B (LVF/CABG) and Group C (MR) had a greater anisotropy index than patients in Group A (CABG) or Group D (AS) at a pacing cycle length of 600 and 400 ms (p < 0.01).
Conduction heterogeneity was assessed during sinus rhythm and pacing from the corners of the plaque. The results are shown in Table 5. Patients in Group B (LVF/CABG) and Group C (MR) had a greater degree of conduction heterogeneity than patients in Group A (CABG). There was no difference in the degree of conduction heterogeneity between Group A (CABG) and Group D (AS). Conduction heterogeneity was not greater when pacing from the inferior plaque corner than when pacing from the superior corner (3.5 ± 1.8 vs. 3.7 ± 2.4; p = 0.33).
Atrial electrogram characteristics
The characteristics of the atrial electrograms were assessed during pacing. In the region of slow conduction, 97.3% of the electrograms were either fractionated or double potentials. In the posterior left atrium, 98.0% of the fractionated signals occurred in the region of slow conduction (Fig. 3). In all patients, the presence of fractionated signals was highly correlated with the region of slow conduction (R = 0.88; p < 0.001).
AERP of the posterior left atrium
There was no significant difference in the AERP at either pacing cycle length between: Group A (CABG) (AERP 600 = 279 ± 25 ms, AERP 400 = 264 ± 24 ms); Group B (LVF/CABG) (AERP 600 = 261 ± 42 ms, AERP 400 = 234 ± 39 ms); and Group D (AS) (AERP 600 = 277 ± 31 ms, AERP 400 = 243 ± 20 ms). The AERP was prolonged in Group C (MR) patients at a cycle length of 600 ms (AERP 600 = 332 ± 42 ms; p < 0.05) compared with the other 3 groups, but this was not significant at a cycle length of 400 ms (AERP 400 = 268 ± 32 ms).
The major findings of this study are as follows. There is a line of functional conduction delay and block in a consistent anatomical location in the PLA running vertically between the pulmonary veins. This line of functional block was seen in all patient groups, but was most marked in conditions associated with greater atrial enlargement (patients with severe mitral regurgitation or severe left ventricular dysfunction). In these patients, there was a greater extent of conduction slowing and block and greater conduction heterogeneity. Conduction in this region showed significant anisotropy with relatively uniform propagation parallel to the line, but marked conduction slowing when wave fronts propagated perpendicular to the line. As a result, wave front propagation took a circuitous course around the region of block establishing the potential for re-entry. During pacing, fractionated electrograms occurred, almost exclusively, along the line of slow conduction.
Anatomically determined functional conduction delay
Animal studies of atrial conduction have shown the development of left atrial conduction slowing in the presence of conditions causing chronic atrial dilatation such as cardiac failure (13,14), mitral regurgitation (15,16), and AV block (17). These conduction abnormalities seemed generalized, and investigators did not emphasize the presence of a critical anatomical region of conduction slowing in either the right or the left atrium. In the sterile pericarditis model, Ortiz et al. (18) showed the importance of a line of functional block in establishing re-entry, the length of the line determining whether re-entrant circuits were stable or unstable. In this study, the line of block had a relatively constant anatomical location.
Prior human studies have evaluated the nature of remodeling in the right atrium in the presence of different forms of structural heart disease resulting in chronic right atrial dilatation. In patients with cardiac failure (19), atrial septal defects (20), or loss of AV synchrony because of asynchronous ventricular pacing (21), there was evidence of both generalized and anatomically determined conduction slowing. The latter invariably involved the region of the crista terminalis where transverse conduction was much more markedly impaired than that in the longitudinal direction, resulting in significant anisotropy. Numerous prior studies have shown the importance of functional conduction block along the crista terminalis in the mechanism of atrial flutter (both typical and atypical) and of AF (22–26). In particular, noncontact mapping studies showed the presence of relatively unstable short cycle length rotors occurring around segments of the crista with slow propagation breaking across 1 or more regions (27,28). Importantly, ablation targeting these regions of conduction slowing was successful in terminating AF.
Markides et al. (8) first showed a similar anatomically determined line of functional block located in the LA of patients with paroxysmal AF. Similar to the crista terminalis in the right atrium, this region corresponded anatomically to a region of change in myocardial fiber orientation. The line descended on the posterior wall of the LA from the roof, passing between the ostia of the superior and then inferior pulmonary veins, in a similar anatomical location to that described in the current study. In that patient population, the line was observed to be responsible for the break up of wave fronts during initiation of AF by pulmonary vein foci. Although the majority of patients in the study by Markides et al. (8) did not have significant structural heart disease or LA enlargement, the presence of AF may promote structural changes (29,30) that contribute to slowing of conduction particularly in anatomical regions predisposed to functional block.
We have extended these observations by showing that this line of block is present in patients with structural heart disease, but is more extensive and associated with greater conduction delay in conditions with greater atrial stretch such as cardiac failure and severe mitral regurgitation. This conduction delay leads to circuitous activation around the entire line of block or can be associated with markedly slowed conduction through a break in the line of block. In either case, this conduction delay creates the substrate for re-entry. Similar to the crista terminalis, this region showed marked conduction anisotropy.
Recent studies have described the targeting of fractionated electrograms from a range of LA sites during ablation of AF (31–33). However, there are conflicting data regarding the utility of this as a primary approach. Importantly, the pathophysiological significance of these signals remains uncertain. In the current study, we observed that fractionated signals were anatomically distributed to the line of slow conduction. Such signals might potentially represent critical zones of slow conduction for short cycle length rotors as described by Lin et al. (27,28) for unstable short cycle length rotors occurring at the crista terminalis. However, it is also probable that many of these signals represent passive slow conduction unrelated to the primary arrhythmia mechanism. Further studies will be required to elucidate this important question.
Prior studies have shown that there may be a discordance between endocardial and epicardial activation in the atrium (34). However, these observations were most pronounced for the trabeculated right atrium, and indeed the posterior wall of the LA seems to be a region where there is concordance between epicardial and endocardial activation (35). The ERP testing was performed at a single site. More sites could not be tested because of the ethical time limitations on the protocol.
There is a line of functional conduction delay in a consistent anatomical location in the PLA of patients with structural heart disease. This is most marked in conditions associated with significant chronic atrial enlargement and leads to circuitous wave front propagation, suggesting a potential role as substrate for atrial arrhythmias.
↵1 Drs. Roberts-Thomson and Haqqani are the recipients of Postgraduate Research Scholarships from the National Health and Medical Research Council (NHMRC) of Australia.
↵2 Dr. Stevenson is the recipient of a Postgraduate Scholarship from the National Heart Foundation of Australia and the Cardiac Society of Australia and New Zealand.
↵3 Dr. Kistler is the recipient of the Neil Hamilton Fairley Scholarship from the NHMRC of Australia and the National Heart Foundation.
This study was supported by a CVL Research Grant.
- Abbreviations and Acronyms
- atrial effective refractory period
- atrial fibrillation
- aortic stenosis
- coronary artery bypass grafting
- left ventricular failure
- mitral regurgitation
- posterior left atrium.
- Received June 18, 2007.
- Revision received October 29, 2007.
- Accepted November 8, 2007.
- American College of Cardiology Foundation
- Kalifa J.,
- Tanaka K.,
- Zaitsev A.V.,
- et al.
- Todd D.M.,
- Skanes A.C.,
- Guiraudon G.,
- et al.
- Morillo C.A.,
- Klein G.J.,
- Jones D.L.,
- Guiraudon C.M.
- Kumagai K.,
- Ogawa M.,
- Noguchi H.,
- Yasuda T.,
- Nakashima H.,
- Saku K.
- Markides V.,
- Schilling R.J.,
- Ho S.Y.,
- Chow A.W.,
- Davies D.W.,
- Peters N.S.
- Lammers W.J.,
- Schalij M.J.,
- Kirchhof C.J.,
- Allessie M.A.
- Li D.,
- Fareh S.,
- Leung T.K.,
- Nattel S.
- Li D.,
- Shinagawa K.,
- Pang L.,
- et al.
- Verheule S.,
- Wilson E.,
- Everett T.,
- Shanbhag S.,
- Golden C.,
- Olgin J.
- Verheule S.,
- Wilson E.,
- Banthia S.,
- et al.
- Neuberger H.R.,
- Schotten U.,
- Blaauw Y.,
- et al.
- Ortiz J.,
- Niwano S.,
- Abe H.,
- Rudy Y.,
- Johnson N.J.,
- Waldo A.L.
- Sanders P.,
- Morton J.B.,
- Davidson N.C.,
- et al.
- Morton J.B.,
- Sanders P.,
- Vohra J.K.,
- et al.
- Sparks P.B.,
- Mond H.G.,
- Vohra J.K.,
- Jayaprakash S.,
- Kalman J.M.
- Olgin J.E.,
- Kalman J.M.,
- Fitzpatrick A.P.,
- Lesh M.D.
- Tomita Y.,
- Matsuo K.,
- Sahadevan J.,
- Khrestian C.M.,
- Waldo A.L.
- Lin Y.J.,
- Tai C.T.,
- Kao T.,
- et al.
- Everett T.H.,
- Li H.,
- Mangrum J.M.,
- et al.
- Ausma J.,
- Wijffels M.,
- Thone F.,
- Wouters L.,
- Allessie M.,
- Borgers M.
- Nademanee K.,
- McKenzie J.,
- Kosar E.,
- et al.
- Oral H.,
- Chugh A.,
- Lemola K.,
- et al.
- Schuessler R.B.,
- Kawamoto T.,
- Hand D.E.,
- et al.