Abstract
Intracardiac echocardiography is not routinely used to guide transseptal puncture (TP) in many centres. TP under fluoroscopy is still the common practice worldwide but remains challenging in difficult cases. This study aims to describe a simple technique to safely localize appropriate TP site during atrial fibrillation (AF) ablation procedure. Inferior vena cava (IVC) angiography was performed at RAO 45°. The IVC, right atrium (RA), right-ventricular inflow tract, and right-ventricular outflow tract were sequentially visualized while the aorta was visualized as non-opacified filling defect. The appropriate TP site was in the middle of the RA, inferoposterior to the non-coronary aortic sinus (NCAS) and superoposterior to coronary sinus ostium. The spatial relationship of these structures was studied in 81 patients. The distance between optimal TP site and surrounding landmarks was analysed. Out of 393 consecutive TPs performed from August 2011 to January 2012, this technique was applied in 17 patients. Under RAO 45° on IVC angiography, an imaginary horizontal line was drawn across the middle point between NCAS and the top of the coronary sinus ostium. The line was divided into four quarters. In 78 (96%) patients, the optimal TP site was identified in the second one. In 94% (16/17) of the patients, all above-mentioned structures were clearly visualized and TP was successfully performed in all of them without complications. IVC angiography is a simple and safe technique which can facilitate TP in difficult cases. Optimal TP site can be easily identified on IVC angiography.
Introduction
Transseptal puncture (TP), a basic technique widely used for variety of percutaneous procedures, has never been more important to electrophysiologists than it is today when therapeutic procedures via a left-atrial (LA) approach, such as atrial fibrillation (AF) ablation and implantation of LA closure device have become daily practice in many electrophysiology (EP) labs.1 In the recent years, intracardiac echocardiography (ICE) has emerged as a useful tool in guiding transseptal left heart catheterization as it provides accurate visualization of the fossa ovalis (FO). However, this imaging modality is not available in many of the EP labs worldwide.1 Traditional TP under fluoroscopy is still the standard practice at many centres around the world, including ours.
Since it was first reported in 1950s, several techniques have been employed to facilitate TP, which increased the success rate and lowered the complication rate of this procedure.1 However, TP remains challenging for inexperienced operators and in difficult cases where distorted atrial anatomy is encountered.2,3
Here we report for the first time inferior vena cava (IVC) angiography-guided TP technique and our experience of a series of patients who had successful IVC angiography-guided TP in the absence of any other imaging modality after failing the conventional technique. Moreover, the spatial relationship between the TP site and surrounding structures is elucidated using computed tomography (CT) and electro-anatomical mapping system (CARTO, Biosense Webster, Inc.).
Methods
Study population
From August 2011 to January 2012, 393 consecutive patients underwent AF ablation at our centre after written informed consent was obtained. During the procedure, TP was first performed using the conventional method (as described in what follows) and the LA was successfully accessed in 376 patients. In the remaining 17 patients (12 men, aged 59.8 ± 8.7 years), TP with the conventional technique failed after at least three attempts by an experienced operator which was defined as an electrophysiologist who had completed at least 1000 TP procedures independently. These 17 patients constituted the population of the present study. In the second part of this study, dual-source CT images were reviewed in another 100 patients who were referred to our centre for AF ablation and the images were utilized for subsequent analysis.
Conventional transseptal technique
A steerable decapolar or fixed quadripolar mapping catheter was positioned within the coronary sinus (CS) through the left femoral vein. The catheter marked both the CS ostium and the inferior margin of the left atrium. An 8-French transseptal sheath and a dilator (SL1, St Jude Medical, MN, USA) were advanced to the superior vena cava (SVC) over a 0.032 inch guide wire inserted via the right femoral vein. The guide wire was removed and the Brockenbrough needle (St Jude Medical, MN, USA) was advanced to within 0.5 cm of the distal tip of the dilator. In the posterior-anterior (PA) fluoroscopic projection, the transseptal assembly was oriented towards the interatrial septum (4 to 5 o'clock position) and was smoothly withdrawn until the second ‘jump’ to the right was visualized, indicating the location of the FO. Once the FO was identified, 45° right anterior oblique (RAO) projection was used. The TP apparatus was slightly rotated anteriorly or posteriorly to let the extremity of the assembly to be vertical to the septum (straightening of the distal curve) and advanced slightly to ensure a good contact with the FO. The needle was then advanced across the septum and LA access was confirmed by contrast injection. After stabilizing the transseptal apparatus inside the LA, the needle was removed and the guide wire was advanced towards the left superior pulmonary vein. If the first attempt failed, the operator would change the puncture site and repeat the TP using the same technique.
Inferior vena cava angiography-guided transseptal puncture
Inferior vena cava angiography was performed in the 17 cases where TP failed at least three times during the initial ablation procedure using the conventional technique. After withdrawing Brockenbrough needle and the dilator, the transseptal sheath was withdrawn 1–2 cm below the IVC ostium. In the 45° RAO projection, 5–10 mL contrast (Iopromide, Ultravist; Bayer Schering Pharma AG, Berlin, Germany) was manually injected through the transseptal sheath and around three cardiac cycle's cine was recorded. The angiography is performed during normal respiration until the IVC ostium, whole right atrium (RA), right-ventricular inflow tract (RVIT), and right-ventricular outflow tract (RVOT) are sequentially and clearly visualized. The non-opacified filling defect between the right atrium and RVOT represents the aortic root and related structures. The angiography images provided information of all the TP-related structures and their proximities including: (i) the angle by which IVC joints RA; (ii) the presence and extent of enlarged IVC and RA dilation; (iii) the posterior, superior, and inferior aspects of the RA wall; (iv) the width and orientation of the ascending aorta; (v) the location of the posterior wall of the aortic root (non-coronary cusp) and tricuspid annular (TA). Based on the angiography images, the optimal TP site was determined as the middle of RA, inferoposterior to the non-coronary aortic sinus (NCAS) and superoposterior to the CS ostium in RAO 45° view (Figure 1).
One case of successful inferior vena cava (IVC) angiography-guided TP after failing conventional method. (A) The Brockenbrough needle could not cross the septum at the site (circle) determined by conventional method in RAO 45°. (B) This figure illustrates IVC angiography on a stop-frame RAO 45° image made after failed TP in (A). The contrast defect region between RA, RVIT, and RVOT is the aortic (AO) root and its adjunctive structures (red dashed contour), and the curve of the CS catheter indicates the top of the CSO (circle). The blue dashed contour lines represent RA. The white arrow indicates the posterior edge of RA and a large IVC is also revealed by angiography. The optimal TPS should be in the middle of the RA. The red circle indicates the failed TP site in (A). (C) The TP site (circle) in (A) is on the posterior edge of RA showed in IVC angiography, it was the reason of a failed TP in (A). (D) After IVC angiography, a puncture is made more anteriorly (circle) than in (A) but in a safer location from AO root. Entry of the needle into the LA is confirmed by contrast injection. (E) The successful TP site guided by IVC angiography. (F) Selective LA angiography in RAO 45° showing the corresponding TPS on LA septum (circle). The blue dashed contour lines represent RA and the white solid contour line indicates LA. The red solid and dashed lines indicate the endocardium and the outer contour of the AO root. Since there is a lot of space between these two lines which also need to be avoided during TP procedure, IVC angiography provides more information than AO root angiography.
Ablation procedure
Our single-catheter technique for AF ablation has been previously described.4 Guided only one transseptal access is normally required through which an ablation catheter is introduced into the LA to generate LA geometry and to perform ablation guided by electroanatomic mapping (CARTO; Biosense Webster). Patients with paroxysmal AF underwent circumferential pulmonary vein isolation (PVI) while in patients with persistent and long-standing persistent AF, linear ablations across the LA roof, mitral isthmus, and cavotricuspid isthmus were subsequently ablated after PVI. Sinus rhythm was achieved by cardioversion if atrial tachycardia or AF persisted. The endpoint of the procedure was PV isolation with complete block of all ablated lines. The ablation catheter was exchanged with a circular mapping catheter that was utilized to verify the isolation of PVs. Conduction block across the lines was validated by pacing manoeuvres.
Computed tomography imaging
To validate our IVC angiography-guided TP technique, we used similar method on reconstructed three dimensional (3D) cardiac shells to localize the optimal TP site. The latter was compared with the real location of the FO identified by CT imaging and the consistency was investigated. From November to December 2011, 100 consecutive patients who underwent an enhanced dual-source CT scanning of heart prior to AF ablation were included in this part of study. CT imaging (Somatom Definition, Siemens, Forchheim, Germany) was acquired 24 h before the procedure and was performed from the level of the ascending aorta to the level of the diaphragm. Sixty millilitres contrast (Ultravist; Bayer Schering Pharma AG, Berlin, Germany) was injected at 4–5 mL/s and image acquisition was taken during suspended breathing and completed within 10 s. Axial images were reconstructed at 0.5 mm intervals at 10% intervals, with the centre of the reconstruction window being between 0 and 90% of one cardiac cycle. At the time of image acquisition, a 70 or 80% phase location which corresponded to atrial end-diastole was selected for image registration in patients with sinus rhythm while a 50% phase location was chosen in those with AF as it yielded the best image quality.
Image segmentation
The CT images were transferred to the electroanatomic mapping system and processed with integrated software (CartoMergeTM Image Integration Module, Biosense Webster Inc.). The techniques used for image segmentation have been previously described.5,6 In brief, the volume of interest was defined first and ‘seeds’ were placed to separate individual anatomic structures. Then the manual annotations were performed on the 3-D CTA data. The volume renderings of IVC, RA, CS, RVIT, RVOT, and aortic root were merged to create a single 3D volume rendering. Opacity of the rendered image of the RA was then reduced to around 80% to make it semitransparent relative to the LA. By this way, we were able to identify the FO and CSO on 3D volume images (Figure 2A). The 5 mm-diameter overlap of RA and LA was defined as the TPS when RA was made semitransparent to LA. Patients were excluded from the study if FO and CSO could not be tagged on the volume-rendered images.
TP site localization on 3D images, drawings, and IVC angiography. (A) Individual cardiac structures are merged into a single image. TPS and CSO are easily identified in the image. This image illustrates the anatomical relationship in Figure 1B. Red chamber represents aorta, purple represents RA, and grey represents LA, light blue represents right ventricular, and bright pink represents pulmonary truck. (B) Diagrammatic representation of the structures visualized in (A). The solid lines indicate the endocardium of each of the structures. The dashed contour line indicates the outer contour of the AO, IVC angiography shows the outer contour of AO, thus help physician avoiding all the tissues between endocardium of RA and AO. So it is much helpful than AO root angiography or positioning of AO pigtail. The superior horizontal line was drawn across the most posterior part of NCAS, the lowest horizontal line was across the top of the CSO, the middle horizontal line was across the middle point of the two lines. The middle line intersects the edge of RA and tricuspid annular at points A and E. The appropriate TPS is from point B to C. (C) This figure illustrates the lines in IVC angiography on a stop-frame RAO 45° image. The blue dashed contour lines represent RA. The red line represents AO root and its surrounding structures. The three horizontal lines were same as those in (B).
Transseptal puncture site localization on 3D images
Under RAO 45° view, an imaginary horizontal line was drawn across the middle point between NCAS and the top of the CSO until it crossed the edge of RA and tricuspid annular (TA) (Line A-E) (Figure 2B). The superior horizontal line was drawn across the most posterior part of NCAS, the lowest horizontal line was across the top of the CSO, the middle horizontal line was across the middle point of the two lines. The line is divided by points B, D, and C into four equal parts. The distance between TP site and the imaginary horizontal line across the top of the CSO was measured, and the distance between TP site and the imaginary vertical lines across points A and E was also measured.
Statistical analysis
Continuous variables are presented as mean ± standard deviation, and categorical variables were shown as percentages or numbers. Intraobserver reproducibility was determined by repeating the measurements at two different time points by one investigator in 20 randomly selected patients. A second investigator performed the measurements in the same 20 patients, providing the interobserver reproducibility data.
Results
The baseline and procedural characteristics of the 17 patients are summarized in Table 1. None of the patients had apparent cardiovascular deformity detected by pre-procedural echocardiography. After multiple failures with conventional TP, single transseptal access was successfully obtained using the IVC angiography-guided technique (15 patients with one attempt and two patients with two attempts). In 16 patients, the appropriate puncture site was inferoposterior to NCAS and superoposterior to CSO. In one patient, the optimal TP site was in the middle of the RA and superoposterior to NCAS and CSO. The following procedure was carried out smoothly without complications and the ablation endpoint was achieved in all cases.
The baseline and procedural characteristics of patients
| Total (n = 17) | |
|---|---|
| Age (years) | 59.8 ± 8.7 |
| Male | 12/17 (71%) |
| AF type | |
| Paroxysmal AF | |
| Persistent or long-standing persistent AF | 29% |
| DM | 1 (6%) |
| HTN | 8/17 (47%) |
| LA diameter (mm) | 37.5 ± 5.3 |
| Left-ventricular EF (%) | 61.5 ± 5.5 |
| Total procedural time (min) | 139 ± 45 |
| Total fluoroscopy time (min) | 29.9 ± 19 |
| Total transseptal puncture time (min) | 4 ± 1.1 |
| Total transseptal puncture fluoroscopy time (min) | 2.3 ± 0.3 |
| Contrast used for IVC angiography (mL) | 6.4 ± 1.9 |
| Total (n = 17) | |
|---|---|
| Age (years) | 59.8 ± 8.7 |
| Male | 12/17 (71%) |
| AF type | |
| Paroxysmal AF | |
| Persistent or long-standing persistent AF | 29% |
| DM | 1 (6%) |
| HTN | 8/17 (47%) |
| LA diameter (mm) | 37.5 ± 5.3 |
| Left-ventricular EF (%) | 61.5 ± 5.5 |
| Total procedural time (min) | 139 ± 45 |
| Total fluoroscopy time (min) | 29.9 ± 19 |
| Total transseptal puncture time (min) | 4 ± 1.1 |
| Total transseptal puncture fluoroscopy time (min) | 2.3 ± 0.3 |
| Contrast used for IVC angiography (mL) | 6.4 ± 1.9 |
LA, left atrial; AF, atrial fibrillation; EF, ejection fraction; IVC, inferior vena cava.
The baseline and procedural characteristics of patients
| Total (n = 17) | |
|---|---|
| Age (years) | 59.8 ± 8.7 |
| Male | 12/17 (71%) |
| AF type | |
| Paroxysmal AF | |
| Persistent or long-standing persistent AF | 29% |
| DM | 1 (6%) |
| HTN | 8/17 (47%) |
| LA diameter (mm) | 37.5 ± 5.3 |
| Left-ventricular EF (%) | 61.5 ± 5.5 |
| Total procedural time (min) | 139 ± 45 |
| Total fluoroscopy time (min) | 29.9 ± 19 |
| Total transseptal puncture time (min) | 4 ± 1.1 |
| Total transseptal puncture fluoroscopy time (min) | 2.3 ± 0.3 |
| Contrast used for IVC angiography (mL) | 6.4 ± 1.9 |
| Total (n = 17) | |
|---|---|
| Age (years) | 59.8 ± 8.7 |
| Male | 12/17 (71%) |
| AF type | |
| Paroxysmal AF | |
| Persistent or long-standing persistent AF | 29% |
| DM | 1 (6%) |
| HTN | 8/17 (47%) |
| LA diameter (mm) | 37.5 ± 5.3 |
| Left-ventricular EF (%) | 61.5 ± 5.5 |
| Total procedural time (min) | 139 ± 45 |
| Total fluoroscopy time (min) | 29.9 ± 19 |
| Total transseptal puncture time (min) | 4 ± 1.1 |
| Total transseptal puncture fluoroscopy time (min) | 2.3 ± 0.3 |
| Contrast used for IVC angiography (mL) | 6.4 ± 1.9 |
LA, left atrial; AF, atrial fibrillation; EF, ejection fraction; IVC, inferior vena cava.
In 94% (16/17) of the patients, all structures related to TP procedure as mentioned earlier were visualized clearly except in one patient where the superior aspect of the RA was not clearly shown.
Out of the 100 individuals, CT image quality was suboptimal and the FO and CSO could not be identified on 3D volume rendering images in 19 patients who were later excluded from the analysis. In the remaining 81 patients, the appropriate TP site was between points B and C in 78 (96%) patients. In these 81 cases, the mean height of the TP site is 0.54 ± 0.09 (the distance between two dashed lines is 1); the mean distance from point A to TP site is 0.47 ± 0.08 (the distance from point A to E is 1) (Figure 2B). In the other three cases, two CT images showed that the TP site was located higher than the NCAS, and was superior-anterior to CSO in one patient.
The average contrast used was 6 mL and X-ray exposure time was 2.3 s for IVC angiogram in these 17 patients.
Follow-up
None of the 17 patients experienced peri-procedural complication. After a mean follow-up of 12 months, four patients had recurrence of AF/atrial tachycardia. One patient underwent a redo procedure during which TP was successful at first attempt with the guidance of an IVC angiography which was recorded 4 months ago during the initial procedure.
Discussion
Locating the posterior and superior boundaries of the RA, AO root, CSO and knowing the proximity of these landmarks are critical when performing a TP. In combination with a CS catheter, an IVC angiogram provides direct visualization of these structures under fluoroscopy by which the location of FO can be easily identified. This technique allows a safe TP at an appropriate site without the need of any additional equipment or facilitates in procedures where LA access is required.
Localizing fossa ovalis and an appropriate puncture site
Originally developed by Cope and Ross et al. in 1959, transseptal technique has become an established technique for obtaining left heart access. Despite the development of new tools for TP such as radiofrequency current, excimer laser, nitinol guidewire, and intracardiac echocardiography,7–11 transseptal access with the Brockenbrough needle under fluoroscopy remains the most common practice. Puncture site and direction is the key to a successful TP procedure. If the needle tip is not placed at an appropriate position and/or is not pointing to the correct direction, damage to adjacent structures can still occur resulting in potentially serious complications. A multi-centre survey showed that the major reason for a failed TP procedure was related to the inability to locate the FO.1 Cardiac perforation complicating with TP is also associated with inadequate localization of the FO.12
Classically, the TP site is determined as one vertebral body above the caudal edge of the LA. However, the direction of the puncture needle tip and the distance from the distal Brockenbrough needle to the posterior LA wall or the mitral annulus are difficult to be precisely assessed. Several techniques have been previously utilized to guide TP. Aortic angiography with a pigtail catheter placed retrogradely at the AO root directly indicates the location of the aorta and aortic valve, thus helps to avoid puncturing into the NCAS.13 However, AO angiography can only outline the endocardial silhouette of the AO (Figure 2C). Actually, the posterior wall of NCAS, the RA septal wall, and surrounding tissues can be found within the limited space between the AO endocardial contour and the RA chamber where the TP needle is placed. But these structures that should be avoided during TP cannot be revealed by AO angiography. Another disadvantage of this technique is that an additional arterial access is required. A His catheter is also used as a reference by some operators. The tip of His catheter indicates the superior margin of the interatrial septum and is adjacent to the NCAS and is the most caudal aspect of the aorta.14 However, the His catheter does not directly mark the AO3 and is prone to displacement during the procedure.
Both ICE and TEE are also helpful in improving the accuracy and safety of TP, however, these techniques require either an extra venous access or a second operator and are not always available in many labs, and last but not least is less cost-effective compared with fluoroscopy-based techniques.15
It was reported that the inferior border of the FO is marked by the IVC with CS at the anteroinferior border and NCAS at anterosuperior border. Cheng et al. reported that the fossa is usually posterior and 1–3 cm inferior to the pigtail catheter placed in the AO, but anterior to the RA silhouette.10 In our study, IVC angiogram could directly or indirectly show all the anatomical references mentioned earlier. The optimal puncture site was easily identified as the point inferior-posterior to the NCAS and superior-posterior to the CSO in RAO 45° projection. This was subsequently validated on 3D CT images. In 78 (96%) patients, the appropriate TPS was to the left of the midline between NCAS and CSO.
Inferior vena cava angiography technique
It is necessary to emphasize that the RAO 45° projection of fluoroscopy is very useful in both choosing an optimal atrial-septal puncture site as well as avoiding puncture of other structures.
Under RAO 45° view, the atrial septum is en face and completely displayed while the intended puncture site is clearly separated from the AO, RA edge, the TA, and the CS. In addition to that, the angle at which the septum is punctured can be visualized. In 1992, Inoue et al.16 recommended using RA angiography to localize TP site under PA view. This RA angiogram does indicate the height of the TP site but is unable to separate the FO from AO root and RA edge.
Our technique of IVC angiography under RAO 45° allows proper visualization of the atrial-septal orientation and anatomic proximity of the surrounding structures, which facilitates a safe and accurate puncture of the interatrial septum. Knowing the exact location of the aortic root, superior and posterior RA wall and superior rim of CSO and the LA inferior border, will create more confidence when performing TP. In three of our cases with remarkable LA enlargement and relatively small RA, the border of posterior RA overlapped with mid-LA under fluoroscopy. There was a possibility that TP using standard technique may puncture RA posterior wall into the pericardial space. This risk can be avoided by IVC angiogram guidance as the RA border was clearly shown.
Placing a deflectable CS catheter via the left femoral vein is the standard practice at our centre. In this way, a CS catheter was always curved to be in contact with the superior boarder of CS and the catheter marked the upper edge of CS ostium under fluoroscopy (Figures 1 and 2). It is suggested that a proper TP site should be superior to the CS ostium because it is usually at the same level as the lower border of the LA near the mitral annular plane. Coronary sinus cannulation through the route of superior vena cava does not possess this advantage.
Validation of TPS determined by inferior vena cava angiogram
Computed tomography, as a non-invasive imaging technique, has become a valuable tool for cardiac chamber segmentation. The method for segmenting the cardiac chambers from CT data sets has shown a high accuracy. The electroanatomic mapping system used in this study has the capability of integrating CT images, which can precisely locate catheter position in relation to real cardiac anatomy and facilitate catheter ablation procedures. The 3D reconstruction of cardiac chambers can provide accurate information about cardiac anatomy. In this study, the 3D images of TP-related structures reveal the relation of cardiac structures in RAO 45° fluoroscopic projection. The proximity of TP site CSO and NCAS measured in 3D CT images is in proportion to that in fluoroscopic view in same projection. Therefore, the feasibility of using IVC angiogram to localize optimal TP site can be validated by the 3D cardiac CT images.
Study limitation
The study population was small. We applied this technique in patients who had failed conventional TP. Whether an IVC angiogram should be routinely performed prior to TP needs to be verified by studies in large cohort with different disease. Furthermore, the LA is much more superior to the RA in emphysema patients or children. In this situation, TP site is not in the middle but superior-posterior of the RA septum wall. And the height for optimal TP site should be determined mainly on the ‘jump’ due to the assembly dropping underneath the superior muscular rim of the FO when the assembly is retracted from the SVC.
Conclusion
Inferior vena cava angiography is a simple and safe technique which facilitates TP by localizing FO and providing the required anatomical information. The optimal TP site on IVC angiogram image is always at the middle part of RA in RAO 45° projection, and inferior-posterior to NCAS and superior-posterior to CSO. This technique is recommended, at least as an adjunctive approach, in complex cases where TP by using conventional technique fails.
Funding
This work was supported in part by grants from Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (ZY201302); China Arrhythmia Registration Study (C-rhythm) from the Ministry of Science and Technology of China (No. 2013BAI09B02).
Conflict of interest: none declared.
References
Author notes
These authors contributed equally to this work, they are co-first authors of this paper.
