© 2005 The European Society of Cardiology. Published by Elsevier Ltd. All rights reserved.
Further evidence of a close anatomical relation between the oesophagus and pulmonary veins
aDepartment of Cardiology and Angiology, University Hospital Münster Albert-Schweitzer-Str. 33, 48149 Münster, Germany; bDepartment of Clinical Radiology, University Hospital Münster Münster, Germany
Manuscript submitted 17 December 2004. Accepted after revision 30 June 2005.
*Corresponding author. Medizinische Klinik und Poliklinik C – Kardiologie und Angiologie, Universitätsklinikum Münster (UKM), D-48129 Münster, Germany. Tel.: +49 251 834 7645; fax: +49 251 834 7864. moennig{at}uni-muenster.de
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
BACKGROUND: Atrio-oesophageal fistula has been reported as a rare but life-threatening complication of ablation of atrial fibrillation (AF). Therefore, the position of the oesophagus in relation to the left atrium (LA) is of major importance for AF ablation.
METHODS AND RESULTS: In order to investigate the possible anatomical variability between the oesophagus and the left atrium, multidetector-row spiral computed tomography (MDCT) of 60 healthy males (age 58.1 ± 5.1 years; LA diameter 5.4 ± 0.7 × 3.8 ± 0.6 cm; LA volume 60.5 ± 15.4 ml) was analyzed. The distance between the oesophagus and the ostia of the pulmonary veins (PV) ranged between 0 and 50.7 mm. Especially for the left PV, the oesophagus was closer than 5 mm to the ostia in 29 cases (48%; n = 24 for left superior PV; n = 10 for left inferior PV; n = 0 for right superior PV; n = 1 for right inferior PV). In addition, the oesophagus was very close to the LA wall (0.8 ± 0.9 mm; range 0–3.3 mm). Intraobserver variability was 1.1 ± 0.7 mm or 3.5%.
CONCLUSION: The position of the oesophagus in relation to the LA and the PV demonstrates high variability. In many cases, the oesophagus is very close to the ostia of the PVs and lies only a short distance from the LA wall. Thus, an anatomical localization of the oesophagus may be critical before or during AF ablation to prevent atrio-oesophageal fistula, especially as there is a need for transmural atrial lesions.
Key Words: atrial fibrillation, atrio-oesophageal fistula, ablation, anatomical position of oesophagus
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
Circumferential pulmonary vein ablation has been reported as an effective preventive therapeutic approach for paroxysmal atrial fibrillation[1–
3]. This catheter intervention is associated with few but potentially life-threatening complications such as oesophageal injury[4,5] and atrio-oesophageal fistula[6]. However, it is unknown whether avoiding ablation at the posterior aspect of the left atrium (LA) is sufficient to minimize the risk of oesophageal complications. We hypothesized that there are important interindividual anatomical variations in the position of the oesophagus in relation to the left atrium and the pulmonary veins which may have important implications for interventional therapy of atrial fibrillation. In order to test this hypothesis, multidetector-row spiral computed tomography (MDCT) was systematically analyzed in a large group of asymptomatic individuals. |
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
Sixty middle-aged asymptomatic male individuals (mean age 58.1 ± 15.2; range 42–64 years) with an increased global risk of myocardial infarction (mean event risk 20% in 10 years calculated by the PROCAM risk score[7]
) but without clinical signs of coronary heart disease were referred for MDCT coronary angiography to determine coronary artery plaque burden and to detect coronary artery lesions. The study was approved by the Institutional Review Board and the patient's written informed consent for the MDCT was obtained.
Scan protocol and image acquisition
MDCT studies were performed on a four-channel multidetector-row CT system (Somatom Volume Zoom®, Siemens Medical Solutions, Forchheim, Germany) using standard parameters for coronary artery CT angiography: detector configuration 4 × 1 mm, 120 kV, effective 400 mA, gantry rotation time 500 ms, table speed 3 mm/rotation. Patients received a beta-adrenoreceptor-antagonist 45 min before the examination (80 mg propranolol orally), if resting heart rate exceeded 60 bpm. A non-ionic contrast material of 140 ml (Iomeprol 300 mg Iodine/ml) was injected via an antecubital vein at 3.5 ml/s followed by a 50-ml saline chaser bolus (LF903000®, Liebel Flarsheim Company, Cincinnati, OH, USA) using a power injector. The scan delay was determined by the test bolus method and based on the peak enhancement at the level of the aortic root.
Image reconstruction
Electrocardiogram-gated image reconstruction was retrospectively performed using the adaptive cardiac volume (ACV) reconstruction algorithm implemented on the scanner (heart rate dependant temporal resolution = 125–250 ms, section thickness = 1.25 mm, increment = 0.8 mm, medium-soft convolution kernel, reconstruction matrix = 512 × 512 pixels, field of view = 180 mm2). The reconstruction window was optimized for a motion free depiction of the coronary arteries performing image reconstruction at midventricular level showing the anterior left ventricular (LV) papillary muscle as well as the anterior and posterior leaflet of the mitral valve in 5%-steps throughout the entire RR-interval.
MDCT data analysis
Reconstructed images were transferred to a 3D-workstation (Leonardo®, Siemens Medical Solutions, Forchheim, Germany) to create multiplanar reformations (MPR) in the axial, the vertical long-axis and horizontal long-axis as well as the short-axis image orientation. Left atrial (LA) and LV volumes were determined using a dedicated software package for cardiac function analysis (CT MASS® 6.1, Leiden, The Netherlands). Endocardial contours were traced automatically including papillary muscles into LV cavity in short-axis as well as axial images with a discernible LV cavity, checked visually for correctness and manually adjusted, if necessary. Maximum LA and LV dimensions were determined from perpendicular measurements in axial and vertical long-axis images. The ostia of the pulmonary veins were identified visually on 2D axial images. The distance between the centre of each PV and the centre of the oesophagus were determined using electronic caliper measurements as provided by the software package (Fig. 1). In addition, the shortest distance between the LA wall and the oesophagus was measured. MDCT data analysis was performed in consensus by two expert readers with a five-year experience in cardiac CT imaging.
|
Statistical analysis
Data are presented as mean ± standard deviation. Univariate analysis was performed by means of the Mann–Whitney-U test for continuous and chi-square test for discrete variables. A value of P < 0.05 was considered as statistically significant. For statistical evaluations, the Statistical Package for the Social Sciences (SPSS for Windows®, Release 9.0) was used.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
All 60 participants were without a history of heart disease and without structural cardiac abnormalities on MDCT. Morphology, diameters and volume of the LA were normal in all individuals (Table 1[8]
). In all participants, four pulmonary veins were present: left superior PV (LSPV), left inferior PV (LIPV), right superior PV (RSPV), and right inferior PV (RIPV). The distance between the inferior ostia (40.1 mm) was slightly longer than between the superior PV ostia (34.5 mm, Table 1).
|
The distance of the oesophagus to each of the pulmonary vein ostia was measurable in all individuals and ranged between 0.0 and 50.7 mm (Fig. 2, Table 2). The shortest distance was the one between the oesophagus and LSPV (6.1 ± 8.8 mm; Table 2, P < 0.001). The distance of the oesophagus to the LIPV was measured as 12.9 ± 13.6 mm. Both RSPV and RIPV are present with the longest distance to the oesophagus (28.6 ± 8.2 and 28.3 ± 8.4 mm, respectively). In 29 cases (48%) the oesophagus was closer than 5 mm to the ostia of the left PVs (n = 24 for LSPV; n = 10 for LIPV; n = 0 for RSPV; n = 1 for RIPV). In nine cases (15%), the oesophagus was to the left of the left PV ostia (n = 6 for LSPV, n = 1 for LIPV, and n = 2 for both). Most cases (n = 45 or 75%) showed an overlap between the oesophagus and the dorsal circumference of the left PV ostia (n = 21 for LSPV, n = 3 for LIPV, and n = 21 for both), whereas only two cases (3%) overlapped the right PVs (n = 1 for RSPV, n = 1 for RIPV; Fig. 2). The oesophagus was parallel to the ostia of the left PVs in 72% of the analyzed cases (n = 43). The distance between the oesophagus and the mid-sternum varied (Table 2). In most cases the oesophagus was to the left but in four cases (7%) it was even to the right of the sternum. In addition, in all individuals the anterior (ventral) aspect of the oesophagus was very close to the LA wall (0.8 ± 0.9 mm; range 0–3.3 mm). In half of the patients (n = 30) there was no fat layer between the oesophagus and the LA at the level of the PVs.
|
|
The intraobserver variability was 1.1 ± 0.7 mm or 3.5% for all measured variables in 20 double analyzed images.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
Catheter ablation has become a new, promising though still developing procedure for treatment of patients with paroxysmal atrial fibrillation. It has been reported that circumferential PV ablation is more effective than segmental ablation of the PV ostia[3]
. Atrio-oesophageal fistula, which may occur as a complication of intraoperative ablation of AF[5], has recently been shown to occur also in percutaneous ablation of the posterior wall of the LA[6,9]. Although the prevalence is very low, it is a very severe complication with two deaths and a residual left hemiparesis in the three reported cases. As there is currently much debate about the need for transmural atrial lesions to achieve freedom from atrial fibrillation recurrence after ablation, the number of affected patients may increase, if deeper atrial lesions are created. In two of these cases, the approximate location of atrio-oesophageal fistula was at the left upper aspect of the posterior LA wall close to the LSPV[6]. This corresponds to our finding that the position of the oesophagus was closest to the LSPV and that there was an overlap between the oesophagus and the LSPV in most of our cases. Thus, one might speculate that RF energy delivery is most dangerous in terms of atrio-oesophageal fistula in the posterior region of the LSPV. A recent anatomical study of left atrial myocardial thickness revealed a variation from 1.2 to 4.4 mm outside the LIPV[10]. Myocardial wall as thin as 1.2 mm might further increase the individual risk of atrio-oesophageal fistula in that region. On the other hand, the anatomical proximity of the oesophagus to the LSPV ostium varies too much to define a safe region for RF delivery. Furthermore, there was no correlation between the site of the oesophagus and the sternum or another fluoroscopically visible structure (e.g. spine) in our cohort. It seems therefore essential to determine the exact oesophageal position in each individual before or during AF ablation in order to avoid RF energy applications in close proximity to the oesophagus. This will most likely reduce the risk of such life-threatening complications.
Recent studies by Pappone et al. and Pachon et al. suggest that complete vagal denervation during circumferential PV ablation or alone reduces the risk of AF recurrence[2,11]. In most of the Pappone groups 102 patients vagal reflexes occurred at the cranial junction of the LSPV and anterior junction of the RSPV during RF delivery. Additionally, such reflexes were also observed at the posteroinferior junction of the LIPV and the RIPV and even at the posterior region of the LA. Therefore, Pappone suggested two additional lines in the posterior LA wall. If information about the exact anatomical localization of the oesophagus is available, a posterior line of ablation may be as safe as a circumferential line in the individual patient. If it is not available, physicians are strongly recommended to avoid a posterior line with high RF energy delivery. In these cases, a roof line in the left atrium may be considered as an alternative, but these patients may not benefit from the vagal denervation of the PVs.
We report the anatomical relationship of the oesophagus to the LA and PV ostia in structurally normal hearts without AF. In most of the patients undergoing AF ablation left atrial dimensions differ considerably from normal. However, our data were concordant in most points with two recent investigations in AF patients[12,13]. The main difference was that we found a closer relation of the oesophagus to the left PVs and an overlap between the oesophagus and PV ostia in several cases. This might be due to the anticipated smaller dimension of the LA in normal hearts without AF.
We found no continuous fat pad between the LA and the oesophagus in 50% of our individuals at the level of PV ostia. This implies that the fat layer might not be the only protector for the oesophagus during AF ablation. Movements and/or contractions of the oesophagus during RF delivery might offer further protection against atrio-oesophageal fistula and might contribute to the relatively low number of reports of this complication in patients undergoing AF ablation. Nonetheless, we found further evidence that the close anatomical relation of the oesophagus and PV ostia has significant implications for AF ablation and we believe that our data justify exact localization of the oesophagus before or during AF ablation.
|
Limitations |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
All CT scans were performed for CHD risk stratification in male, healthy individuals and retrospectively analyzed for the anatomical relation between PVs and oesophagus. But PV anatomy was suitable for analysis in all 60 CT scans. Although none of the subjects in this study underwent AF ablation, the data fit well with other cohorts of AF patients[12,
13]. Furthermore, this patient group can serve as controls presenting electrically and anatomically "normal" hearts. |
Conclusion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
Our data show that the normal anatomical position of the oesophagus in relation to the LA and the PV demonstrates high variability. In most cases, the oesophagus is very close to the ostia of the left PVs and is only a short distance away from the LA wall. Thus, anatomical localization of the oesophagus may be critical before or during AF ablation to prevent atrio-oesophageal fistula, especially in view of a need for transmural atrial lesions.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
We thank Stephanie Fockenberg and Martin R. Fuchs for excellent technical assistance.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionLimitationsConclusionAcknowledgementsReferences |
|---|
[1] Pappone C., Rosanio S., Augello G., et al. Mortality, morbidity, and quality of life after circumferential pulmonary vein ablation for atrial fibrillation: outcomes from a controlled nonrandomized long-term study. J Am Coll Cardiol 2003; 42: 185–197.
[2] Pappone C., Santinelli V., Manguso F., et al. Pulmonary vein denervation enhances long-term benefit after circumferential ablation for paroxysmal atrial fibrillation. Circulation 2004; 109: 327–334.
[3] Oral H., Scharf C., Chugh A., et al. Catheter ablation for paroxysmal atrial fibrillation: segmental pulmonary vein ostial ablation versus left atrial ablation. Circulation 2003; 108: 2355–2360.
[4] Gillinov A.M., Pettersson G., Rice T.W. Esophageal injury during radiofrequency ablation for atrial fibrillation. J Thorac Cardiovasc Surg 2001; 122: 1239–1240.
[5] Doll N., Borger M.A., Fabricius A., et al. Esophageal perforation during left atrial radiofrequency ablation: is the risk too high? J Thorac Cardiovasc Surg 2003; 125: 836–842.
[6] Pappone C., Oral H., Santinelli V., et al. Atrio-esophageal fistula as a complication of percutaneous transcatheter ablation of atrial fibrillation. Circulation 2004; 109: 2724–2726.
[7] Assmann G., Cullen P., Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Münster (PROCAM) study. Circulation 2002; 105: 310–315.
[8] Juergens K.U., Schulze Eilfing B., DeAngelis G., et al. Anatomy and morphology of the heart based on multidetector-row computed tomography coronary angiography: morphometric analysis of 60 middle aged asymptomatic male individuals. Eur Radiol 2004; 14: R3.
[9] Scanavacca M.I., D'avila A., Parga J., Sosa E. Left atrial-esophageal fistula following radiofrequency catheter ablation of atrial fibrillation. J Cardiovasc Electrophysiol 2004; 15: 960–962.[Web of Science][Medline]
[10] Becker A.E. Left atrial isthmus: anatomic aspects relevant for linear catheter ablation procedures in humans. J Cardiovasc Electrophysiol 2004; 15: 809–812.[CrossRef][Web of Science][Medline]
[11] Pachon M.J.C., Pachon M.E.I., Pachon M.J.C., et al. A new treatment for atrial fibrillation based on spectral analysis to guide the catheter RF-ablation. Europace 2004; 6: 590–601.
[12] Lemola K., Sneider M., Desjardins B., et al. Computed tomographic analysis of the anatomy of the left atrium and the esophagus. Implications for left atrial catheter ablation. Circulation 2004; 110: 3655–3660.
[13] Kottkamp H., Piorkowski C., Tanner H., Kobza R., Dorszewski A., Schirdewahn P., et al. Topographic variability of the esophageal left atrial relation influencing ablation lines in patients with atrial fibrillation. J Cardiovasc Electrophysiol 2005; 16: 146–150.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. H. Hauser, D. C. Peters, J. V. Wylie, and W. J. Manning Evaluating the left atrium by magnetic resonance imaging Europace, November 1, 2008; 10(suppl_3): iii22 - iii27. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Knackstedt, P. Schauerte, and P. Kirchhof Electro-anatomic mapping systems in arrhythmias Europace, November 1, 2008; 10(suppl_3): iii28 - iii34. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kobza, T. Treumann, and P. Erne Visualization of the oesophagus in relation to the left atrium: an alternative concept Europace, January 1, 2007; 9(1): 64 - 65. [Full Text] [PDF]
|
|
|
|
|
|
B. Nilsson, X. Chen, S. Pehrson, and J. H. Svendsen The effectiveness of a high output/short duration radiofrequency current application technique in segmental pulmonary vein isolation for atrial fibrillation. Europace, November 1, 2006; 8(11): 962 - 965. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Gula, A. C. Skanes, E. Posan, A. D. Krahn, R. Yee, and G. J. Klein Gastroesophageal Reflux Facilitates Esophageal Imaging During Pulmonary Vein Ablation Circulation, August 8, 2006; 114(6): e235 - e236. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||