Vol.48 - Número 1, Enero/Marzo 2019 Imprimir sólo la columna central

Usefulness of magnetic resonance in the analysis of the anatomy of the pulmonary
veins in patients with atrial fibrillation waiting for radiofrequency ablation


Instituto Modelo de Cardiología de Córdoba S.R.L.
(5003) Córdoba, Argentina
Recibido 29-SET-18 – ACEPTADO después de revisión el 29-NOV-2018.
There are no conflicts of interest to disclose.



Nowadays, pulmonary veins (PV) ablation in patients with symptomatic drugs refractory atrial fibrillation (AF) has become the treatment of choice. Objective: To determine the usefulness of Magnetic Resonance (MR) of left atrial and pulmonary veins as reference anatomic parameter for navigation systems in patients who are going to be submitted to AF radiofrequency ablation.
Material and Methods: This is a retrospective case control study, which included 91 consecutive patients with AF diagnosis waiting for ablation between July 2014 and December 2017. The studied population was a mean 58.2 years old (range: 21-82 years old), and 72.5% were men. MR and the EnSite navigation system imaging were compared. Patients with the presence of left atrial/appendage clot were excluded.
Results: Our studied population presented more frequently an anatomic pattern of four PV. Supernumerary PV was the anatomic variant found in the majority of cases, with all patients showing a right intermediate vein. The common branch/ostium was the less usual variant, and it was left sided in all cases. The MR pulmonary veins anatomic analysis showed a substantial concordance with the EnSite navigation system, making the radiofrequency ablation planning easier.
Conclusions: MR is an efficient, safe and simple tool which is highly precise in the evaluation of the anatomy of the left atrium and the PVs, so that its integration to navigation systems should be disseminated.
Key words: Atrial fibrillation. Ablation. Magnetic resonance.


Atrial fibrillation (AF) ablation is the treatment of choice for those patients that remain symptomatic in spite of the pharmacological antiarrhythmic treatment [1,2]. The main goal in AF ablation is to achieve the complete electrical isolation of pulmonary veins (PVs).

The typical anatomical pattern of PVs is the presence of four vessels of drainage independent from the LA; two right ones and two left ones. There are normal anatomical variants of venous drainage into the LA that have been proven by different studies [3,4].

Magnetic resonance imaging (MRI) provides images with a great time-space resolution, being a very useful imaging method to make the anatomical and morphometric study of PVs and the left atrium (LA); and also, it provides information on the existence of intracavitary thrombi, which are usually lodged in the left atrial appendix in this pathology.

The aim of this study is to determine the usefulness of MRI of the LA and PVs as a reference anatomical parameter for the navigation systems in patients waiting for AF ablation.



  • POPULATION. This is an observational retrospective study, of the case-control type, where 91 consecutive patients with diagnosis of AF refractory to pharmacological treatment, waiting for radiofrequency (RF) ablation, were studied. 92.9% were studied with MRI within the 48 h before the procedure, in a term from July 2014 to December 2017.
  • NUCLEAR MAGNETIC RESONANCE. The anatomy of PVs, LA and their anatomical variants were assessed by MRI. A 1.5 TESLA scanner (SIEMENS MAGNETOM ESSENZA) was used to perform the chest magnetic resonance angiography, in some cases associated to cardiac MRI, with specific coil of sixteen channels, electrocardiographic and respiratory synchronization. The studies were made obtaining images during apneas requested to the patient. The sequences used in the MR angiography protocol were T2 TRUFI and T1 black-blood in coronal, sagittal and axial planes, axial T1 VIBE with no contrast, MR angiography sequences with paramagnetic contrast in coronal plane in arterial and venous phase of 1 mm of thickness with posterior axial isotropic MIP reconstruction and coronal of the left atrium and pulmonary veins, performing 3D reconstructions and final T1 VIBE axial sequence with contrast. Cardiac MRI comprised morphological axial sequences T1-T2 at the level of the cardiac area and supraaortic trunks, cine MRI sequences in cardiac longitudinal axes of two chambers, four chambers and three chambers, short axis, flow velocimetry and late enhancement sequences.
    Axial and coronal isotropic MIP reconstructions of MRI were transferred to the electroanatomical mapping system, 3D EnSite, obtaining a 3D reconstruction of LA and the proximal portion of PVs.
  • RADIOFREQUENCY ABLATION. RF ablation procedure was made under general anesthesia. Right femoral venous puncture was made by the modified Seldinger technique, introducing three catheters, a quadripolar one for the coronary sinus and through transeptal puncture guided by contrast, a duodecapolar spiral catheter was made to advance into the LA to record the activity of the PVs and a 4 mm catheter with irrigated tip for mapping and ablation.
    Through the EnSite velocity electroanatomical navigation system (advanced cartography and navigation system for catheters), LA reconstruction was made, which was subsequently fused to MRI images. The exclusion criteria were the presence of thrombus in the atria and/or left appendices and patients younger than 18 years.
  • STATISTICAL ANALYSIS. The data were extracted from the analysis of clinical histories of patients admitted during the described period, being gathered with the Excel, Microsoft® software. Later, a database was generated in SPSS, IBM® for the statistical analysis.
    The numerical variables were presented as medians and interquartiles and means and standard deviations and the nominal variables in percentages. The comparisons of the numerical variables were made with the Student’s t test and if the distribution was abnormal, the Wilcoxon test was applied; the nominal variables were analyzed with the Chi-square or Fisher’s tests as it corresponded.
    Kappa coefficient was conducted to observe the reliability of MRI to properly identify each of the veins, including the anatomical variables and to compare them to the navigation system. To interpret the reliability of the interpretation between observers, the following qualitative terms were applied: 0-0.2 (mild); 0.2-0.4 (just); 0.4-0.6 (moderate); 0.6-0.8 (substantial); 0-8-1 (perfect) [5].


The variables studied were age, gender and number of PVs. Age was taken in years and to represent it, medians and means were used; and gender was represented as a nominal binary variable.

The number of PVs with outlet into the LA was counted. The typical anatomy of four PVs was the most usual finding. This corresponds to two superior veins, right and left, where the left one usually has a more superior outlet in comparison to the right one, and the projection of both is normally forward and upward; and two inferior veins, right and left, with a more posterior outlet with a backward and downward projection. Also, the existence of anatomical variants described in the bibliography were analyzed (Figure 1).

Figure 1. Anatomical patterns of pulmonary veins (PVs) anatomy in patients with AF and controls. The gray portions indicate variants of typical anatomy. A) Typical pattern of PV; B) left common ostium or short trunk pattern; C) left common trunk pattern; D) intermediate right PV; E) two intermediate right PV; F) intermediate right PV and superior right PV. (Modified from: Ritsuchi Kato et al. Pulmonary Veins Anatomy in Patients Undergoing Catheter Ablation of Atrial Fibrillation. Circulation 2003) [2].


The anatomy of the ostium is considered the area of coalescence of the PV wall with the LA wall. Taking into account this description, common ostium was considered the coalescence of the superior PV inferior wall and the inferior PV superior wall outside the LA annulus. In the case there was a branching of at least 5 mm between this common ostium and the LA, it was considered as the anatomical variable called common trunk. Figures 2, 3, 4 and 5 show different anatomies of the PVs found in the studied population.

Figure 3. 3D magnetic resonance imaging of LA and PV. It shows PV and a fifth right intermediate vein of small caliber, which appears ahead of the right inferior PV.

Figure 4. Magnetic resonance imaging: coronal MIP of LA and PV. It shows left common trunk (arrow) and right superior PV. The right inferior PV is not observed because it is in a more inferior-posterior plane.

Figure 5. Reconstruction of the LA and PV with navigator (3D gray image), being supported on the anatomical parameter of the magnetic resonance angiography (3D red image).

There were 91 consecutive patients studied who consulted to undergo AF RF ablation. The average age was 58.2 years. The male gender was the predominant one in patients with arrhythmias, representing 72.5% of cases, while 27.5% were females. MRI was carried out in 92.9% of patients. The remaining 7.1% had devices not compatible with MRI (pacemakers/implantable defibrillators) and had been studied by transesophageal echo in other centers (Table 1).

Table 1. Characteristics of the population studied
  Age (years) 58.2 (21-82)
  Gender (%)
• Female
•  Male

  Magnetic resonance (%) 92.9%
  Age (years) 58.2 (21-82)

The anatomical characteristics of PVs found showed a typical pattern of four veins in 83.6% of patients; while 16.4% presented variants. From these variants the presence of 5 PVs was observed in 9 patients (9.8%) and the common ostium/trunk variant in 6 patients (6.6%) of the total sample. The whole variant of 5 PVs corresponded to a fifth right intermediate vein. All common ostium/trunks are found at the left side. Table 2 summarizes the findings found in our studied population.

Table 2 . Anatomical characteristics of PVs
  PVs Population (%)
  Anatomical pattern of PVs
- Typical (four veins)
- Variant

  Intermediate PV
- Right
  Common ostium/trunk
- Left
- Right


In the analysis of anatomical variants outlined in relation to gender, it was observed that 93.3% (14 patients) was present in male patients; while only 6.6% (1 patient) was female. The presence of a fifth PV in all cases was in men; while the common trunk/ostium presented in 83.3% of men.

By concordance analysis, using the Kappa score, to learn the reliability of MRI and the navigation system in the analysis of the PVs anatomy and its variants, a comparison about the capacity of analysis of each of these methods was conducted, finding that there is a level of substantial concordance (0.47±0.22, confidence interval of 95%) (Figure 5 shows LA and PV fusion with MRI and navigator).


AF is the most prevalent sustained arrhythmia in the general population, with a high rate of morbidity and mortality associated to embolic phenomena, deterioration of functional class and consultations to the ER.

Its mechanisms are due to active ectopic foci that are nearby the atrial insertion of the pulmonary veins [6,7,8,9].

In prospective studies, the incidence of AF increases 0.1% per year in patients younger than 40 years, at 1.5% per year in women and 2% per year in men older than 80 years. The risk of suffering atrial fibrillation having turned 40 is approximately 25%. This makes it an expensive disease, triggering the search for strategies to decrease its negative impact on the health of individuals and its cost for society.

This study yielded results similar to those of literature, taking into account the average age (58.2 years) of patients with AF in the studied population. Also, there was a higher prevalence of the male gender as described in literature [8,9,10].

In this study, 16.4% of patients presented a PV anatomy different from that considered typical of 4 vessels, coinciding with other studies as to the anatomical variants [7,8,9].

In this work, the presence of supernumerary veins prevailed on the right side, and in all cases it corresponded to an intermediate PV with origin in the middle lobe, just as in other studies [7,8,9].

The presence of left common trunk and ostium was less frequent (6.6%) in comparison to other studies, which showed a prevalence close to 33.7% [8,9,10]. The low percentage of 6.6% could be explained due to the small size of the population sample. We should highlight that variations according to race and associated co-morbidities were not taken into account.

The anatomical information contributed by MRI was fused with LA and the proximal portion of the PV mapping made with the EnSite navigation system during ablation, showing a substantial concordance between both methods (Figures 5 and 6).

Figure 6. Fusion of images between the navigator and the magnetic resonance angiography of LA and PV.


This study showed the MRI is a very useful imaging technique due to the accurate anatomical reproduction of the PV and the LA, not requiring exposing the patients to ionizing radiation and iodinated contrast used by other methods, such as CT.

Another very valuable piece of information contributed by MRI is its high sensitivity and specificity to detect thrombi in the left atrium and appendix, not requiring transesophageal echo (Figures 7 and 8).

The detection of anatomical variants of PV and LA before ablation simplifies and makes the procedure more dynamic.

Figure 7. SSFP sequences, vertical longitudinal axis of the left ventricle, where hypointense rounded image is observed, adhered to the appendix roof, which corresponds to thrombus.

Figure 8. Magnetic resonance angiography with isotropic axial reconstruction of appendix, which is free from thrombi.


The most important ones come from the retrospective analysis and the small number of patients included in the study.


MRI enables an excellent anatomical correlation of the LA and the PV when integrating them to the navigation system. It prevents the exposition of the patient to ionizing radiations and iodinated contrasts, and it allows ruling out with a very high accuracy, the presence of intracavitary thrombi, so transesophageal echo is not necessary before ablation.

In the series analyzed here, the incidence of anatomical variants was greater in the male gender (21.2% of the total of 66 men); while in the female gender, these variants were observed in only 4% of the total of 25 women.



  1. Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC Guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J 2016; 37 (38): 2893-2962.
  2. Andrade JG, Macle L, Nattel S, et al. Contemporary atrial fibrillation management: A comparison of the current AHA/ACC/HRS, CCS, and ESC Guidelines. Can J Cardiol 2017; 33 (8): 965-76.
  3. Ritsuchi K, Lickfett L, Meininger G, et al. Pulmonary anatomy in patients undergoing catheter ablation of atrial fibrillation. Circulation 2003; 107 (15): 2004-10.
  4. Thorning C, Hamady M, Liaw JV, et al. CT evaluation of pulmonary venous anatomy variation in patients undergoing catheter ablation for atrial fibrillation. Clin Imaging 2011; 35 (1): 1-9.
  5. Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical epidemiology. A basic science for clinical medicine, 2nd edn. Little Brown, Boston, 1991.
  6. Stewart S, Hart CL, Hole DJ, McMurray JJ. Population prevalence, incidence, and predictors of atrial fibrillation in the Renfrew / Paisley study. Heart 2001; 86 (5):516-21.
  7. Reinhold, Lindig C, Willich SN, Brüggenjürgen B. The costs of atrial fibrillation in patients with cardiovascular comorbidities a longitudinal analysis of German health insurance data. Europace 2011; 13 (9): 1275-80.
  8. Piccini JP, Lopes RD, Kong MH, et al. Pulmonary vein isolation for the maintenance of sinus rhythm in patients with atrial fibrillation: A meta-analysis of randomized, controlled trials. Circ Arrhythm Electrophysiol. 2009; 2 (6): 626-33.
  9. Rivard L, Hocini M, Rostock T, et al. Improved outcome following restoration of sinus rhythm prior to catheter ablation of persistent atrial fibrillation: A comparative multicenter study. Heart Rhythm. 2012; 9 (7): 1025-30.
  10. Lemery R. Catheter ablation of paroxysmal atrial fibrillation: long-term follow-up and the inevitability to fibrillate. Europace. 2011; 13 (3): 301-3.

Publication: March 2019


Revista de FAC


Contenidos Científicos
y Académicos