Abstract
A detailed understanding of the anatomy of the atrial septum and its relationship with critical structures such as the aortic root and the posterior atrial wall is crucial to safely perform transseptal catheterization. A biplane fluoroscopy-guided technique using conventional fluoroscopic landmarks guided by diagnostic catheters in standard positions (e.g., coronary sinus, His bundle, noncoronary sinus of Valsalva) is the historical standard for atrial transseptal catheterization. Intracardiac echocardiography and other specialized tools, such as radiofrequency-assisted transseptal needles, have significantly improved the efficacy and safety of atrial transseptal catheterization. These tools greatly facilitate transseptal catheterization in patients with challenging anatomy, such as a thickened or aneurysmal interatrial septum or the presence of an atrial septal closure device.
Keywords
device transseptal catheterization, fossa ovalis, interatrial septal occlusion, intracardiac echocardiography, left atrial appendage, occlusion device, transesophageal echocardiography
Key Points
- •
A detailed understanding of the anatomy of the atrial septum and its relationship with critical structures such as the aortic root and the posterior atrial wall is crucial to safely perform transseptal catheterization.
- •
A biplane fluoroscopy-guided technique using conventional fluoroscopic landmarks guided by diagnostic catheters in standard positions (e.g., coronary sinus, His bundle, noncoronary sinus of Valsalva) is the historical standard for atrial transseptal catheterization.
- •
Intracardiac echocardiography and other specialized tools, such as radiofrequency-assisted transseptal needles, have significantly improved the efficacy and safety of atrial transseptal catheterization. These tools greatly facilitate transseptal catheterization in patients with challenging anatomy, such as a thickened or aneurysmal interatrial septum or the presence of an atrial septal closure device.
Acknowledgement
The authors would like to thank Michael Seckeler, MD, MSc, FACC for his assistance in editing the section pertaining to transseptal catheterization in patients with congenital heart disease.
Introduction
Access to the left heart via the atrial transseptal catheterization (TSC) was first described in 1959 by Ross and colleagues as an alternative technique to the conventional transaortic approach to obtain left-sided hemodynamic measurements. Notwithstanding subsequent technical refinements pioneered by Braunwald, Brockenbrough, and Mullins, TSC for hemodynamic studies was initially adopted only by few highly specialized institutions, because of the higher risk of potentially life-threatening complications such as cardiac perforation. The development and widespread implementation of radiofrequency (RF) catheter ablation for the treatment of cardiac arrhythmias in the 1990s renewed interest in the transseptal approach to gain access to left heart chambers for mapping and ablation. After the pivotal demonstration by that focal discharges from the pulmonary veins (PVs) are implicated in the initiation of human atrial fibrillation (AF), TSC became critical for catheter-based procedures to eliminate arrhythmogenic triggers from the PVs, and is now considered a fundamental skill for any interventional electrophysiologist. Although the general technique of TSC has not changed substantially over the years, important technologic advances have made the procedure significantly easier and safer. This chapter will review the indications, techniques, and outcomes for atrial TSC for invasive electrophysiologic procedures.
Anatomic Considerations for Transseptal Catheterization
A detailed understanding of the anatomy of the atrial septum is crucial to safely perform TSC. When the heart is viewed from an attitudinal perspective, the right atrium (RA) is rightward and anterior, whereas the left atrium (LA) is a leftward and posterior structure. As a result, the plane of the interatrial septum (IAS) is not described by the anteroposterior sagittal plane, but is instead slanted from left anterior to right posterior. Fluoroscopically, the IAS is almost perpendicular to the plane of the screen in the left anterior oblique (LAO) projection and faces the plane of the screen in the right anterior oblique (RAO) projection. Although the septal RA and LA walls are large structures, the true septum suitable for TSC is considerably smaller and coincides with the fossa ovalis and its limbus or muscular rim. In a transesophageal echocardiography (TEE) study, Schwinger et al. showed that the muscular rim around the fossa ovalis is not always a distinct and prominent structure, and a gradual thinning without a clear muscular rim can be found in up to 20% of cases ( Fig. 39.1 ). This is particularly important for the TSC technique, because most operators rely on the tactile and visual feedback of a “jump” when withdrawing the transseptal sheath and needle from the superior vena cava (SVC) to the fossa ovalis, corresponding to the passage between the limbus (superior) and the fossa ovalis (more inferior) (see Section 5.2). It is important to emphasize, to avoid complications such as intramural septal hematoma, particularly in fully anticoagulated patients. That although the muscular rim is a part of the true septum, it should not be routinely targeted for TSC ( Fig. 39.2 , 
). The fossa ovalis, the only structure that should be targeted for TSC, is a fibromembranous structure with a thickness ranging from 0.5 to 1.5 mm. When viewed from within the RA, the fossa ovalis appears as a crater-like translucent depression. The normal fossa ovalis is an oblong structure; its superior-inferior diameter ranges from 10 to 31 mm, whereas the anterior-posterior diameter measures 5 to 14 mm. The larger superior-inferior diameter of the fossa ovalis has relevant implications when double TSC is required. In these cases, it may be easier and safer to accommodate the transseptal sheaths obtaining accesses at different “heights” (one more superior and one more inferior), rather than at different planes in the anterior-posterior dimension (see Fig. 39.1 ).
The RA septum located anterior to the fossa ovalis (which extends to the septal leaflet of the tricuspid valve) overlies the transverse pericardial sinus and the aortic root; this invaginates the RA at the level of the noncoronary sinus of Valsalva. The RA septal wall posterior to the fossa ovalis is in continuity with the pericardial space. Therefore puncturing the RA septum outside the fossa ovalis is associated with high risk of cardiac perforation and/or puncture of the aortic root, with potential for catastrophic complications (see Fig. 39.1 ).
Indications for Atrial Transseptal Catheterization
The location of septal crossing can be optimized for the specific procedure performed ( Fig. 39.3 , 
). In general, a posterior crossing is optimal when targeting posterior LA structures (e.g., the PVs during AF ablation). For patients undergoing AF ablation with either magnetic navigation or balloon technologies, it may be more favorable to puncture the fossa in a more anterior and inferior location. The magnetic navigation catheter requires a greater working length within the LA to allow full deployment of the catheter-based magnets. An anterior and inferior fossa approach for balloon PV ablation greatly facilitates access to the right inferior PV. An anterior approach is also more favorable to access the left atrial appendage (LAA) (e.g., to place percutaneous LAA closure devices) or the mitral valve annulus (e.g., to gain access to the left ventricle (LV) for mapping and/or ablation, for placement of endocardial pacing leads, or to target a left-sided accessory pathway).
A transseptal approach is also recommended when there are contraindications to retrograde transaortic approach to the LV, such as the presence of severe aortic valve disease, a mechanical aortic prosthesis, significant aortic atherosclerosis, and/or an aortic aneurysm. For all the other left-sided procedures, such as mapping and ablation of accessory pathways or ventricular tachycardia (VT), TSC is alternative to the conventional retrograde transaortic approach, with different relative merits and limitations. For instance, the TSC has the advantage of avoiding arterial access, thus minimizing the time needed for complete postprocedural vascular recovery. In our experience, mapping the circumference of the mitral annulus is more easily performed via a transseptal approach; this is particularly evident during catheter mapping of left-sided accessory pathways. These two approaches were compared in a series of 106 patients undergoing catheter ablation of a left-sided accessory pathway. A transseptal approach was adopted as a first-line method in 51 (48%) subjects; the remaining patients underwent ablation with a conventional retrograde approach. The authors reported no difference in total procedure time (220 ± 12.8 min vs. 205 ± 12.5 min) or fluoroscopy time (44.1 ± 4.4 min vs. 44.7 ± 5.1 min). Of note, the retrograde approach was associated with higher incidence of periprocedural complications or crossover to the other technique (42% vs. 11%, P <.01). Similar results have also been reported in another series. Although the retrograde approach is often preferred to TSC for mapping and ablation of VT, certain structures (e.g., papillary muscles) may be more easily sampled with a TSC approach in selected patients. The presence of left atrial thrombus or mobile mass constitutes a relative contraindication to TSC. In addition, TSC should be avoided for patients in whom persistent right-to-left shunting would be unfavorable (e.g., concomitant LV assist device).
Intraprocedural Patient Management: Sedation, Anticoagulation Status
Proper procedural sedation is important to avoid unpredictable patient movements and/or respiratory excursions during the TSC, which might result in significant shifts of the transseptal sheath and needle positioning once the fossa ovalis is engaged. A recent study evaluated the impact of the phase of respiration on catheters positioned in the central venous system using computed tomography (CT). This study showed that, during inspiration, a catheter positioned in the RA might shift superiorly by an average of 9 mm. Such an inspiratory shift during TSC could result in inadvertent puncture of the muscular rim of the septum or the roof of the LA.
The practice of uninterrupted warfarin during AF ablation is commonly used. A recent metaanalysis including more than 27,000 patients undergoing catheter ablation of AF showed a significant reduction in periprocedural thromboembolism with an uninterrupted warfarin strategy compared with low-molecular-weight heparin bridging (odds ratio [OR], 0.10; 95% confidence interval [CI], 0.05 to 0.23; P <.001). Although the incidence of major bleeding complications did not differ between the two anticoagulation strategies, minor bleeding was significantly less common in patients undergoing ablation procedures during uninterrupted warfarin therapy (OR 0.38; 95% CI, 0.21–0.71; P =.002).
Before the widespread implementation of intraprocedural imaging with intracardiac echocardiography (ICE) to guide the TSC, systemic anticoagulation was typically withheld until the achievement of LA access to minimize the risk of major bleeding (e.g., in case of inadvertent puncture of the aorta or cardiac perforation). Delaying the initiation of systemic anticoagulation until after TSC may increase the risk of sheath-associated thrombus formation detected with ICE imaging; thus we routinely initiate anticoagulation before insertion of the transseptal sheaths ( Fig. 39.4 , 
). In a subgroup analysis of eight studies included in the metaanalysis by Santangeli et al., four prescribed heparin administration before TSC. The composite end point of major bleeding and periprocedural systemic thromboembolism occurred in 75/4257 (1.76%) patients in whom heparin was administered before left atrial access, as compared with 16/436 (3.67%) of those in whom heparin was administered immediately after the transseptal access (TSA) (OR, 0.47; 98% CI, 0.27–0.81; P = .007). These results strongly support the administration of heparin before left atrial access, particularly when ICE imaging is available to optimize the puncture site. In addition, lower levels of anticoagulation (activated clotting times 250–300 seconds) are associated with an increased incidence of both sheath-associated and in situ thrombosis in patients undergoing AF ablation; using a higher target activated clotting time (ACT; 300–350 seconds) may decrease this occurrence.
Techniques and Tools for Transseptal Catheterization
General Considerations
The basic tools for TSC are the transseptal sheath and needle ( Fig. 39.5 ). Standard nonsteerable transseptal sheaths (e.g., LAMP series or SL series, St. Jude Medical, St. Paul, MN) are available in two sizes (8 F and 8.5 F) and different lengths ranging from 63 to 81 cm with built-in 67 to 85 cm dilators; the system can accommodate a 71- to 89-cm Brockenbrough needle or a standard 0.032-inch guidewire. Steerable sheaths (e.g., Agilis™, St. Jude Medical, St. Paul, MN) are usually available in only one size (8.5 F) and two lengths (61 cm or 71 cm), and can also accommodate a 71- to 89-cm transseptal needle or a 0.032-inch guidewire.
The design of the transseptal needles is standardized and consists of a proximal hub containing a flange with a flat and a pointed end ( arrow , indicating the plane of the needle curvature) and a distal end. The great majority of transseptal punctures are performed with the standard needle curvature (BRK, St. Jude Medical, St. Paul, MN), although different curvatures might be needed for large (BRK-1) or very small (BRK-2) atria. The choice of transseptal sheath and needle can be tailored to the operator’s preference and/or individual patient anatomy; however, for AF ablation procedures, the use of steerable sheaths has been shown to improve procedural outcomes. A recent prospective randomized trial evaluated the benefit of a steerable sheath compared with a conventional nonsteerable one. Freedom from recurrent AF/atrial tachycardia after a single procedure was significantly higher in patients undergoing ablation with a steerable sheath (76% vs. 53% after 6 months, P = .008); in addition, fluoroscopy time was lower when a steerable sheath was used (33 ± 14 min vs. 45 ± 17 min, P < .001). In a recent study by our group including 300 consecutive patients undergoing AF ablation, the use of steerable sheaths was also shown to improve long-term arrhythmia-free survival as well as rates of both acute and chronic PV reconnection. Therefore the available evidence largely supports the use of steerable sheaths in the setting of AF ablation. The use of a larger diameter curve for the steerable sheath is quite useful in AF ablation patients with dilated atria or when ablation at the valve annuli is required (e.g., mitral or tricuspid flutter ablation, LV ablation). In the following section, the specific techniques to obtain TSC will be reviewed.
Fluoroscopy-Guided Transseptal Catheterization
The fluoroscopy-guided technique is still commonly used for atrial TSC. A correct understanding of the fluoroscopic anatomy based on conventional landmarks is crucial to safely perform the TSC ( Fig. 39.6 ). To locate the approximate region of the fossa ovalis, it is important to validate the position of the IAS and of other critical anatomic structures, such as the aortic root. For this purpose, diagnostic catheters are usually positioned in standard locations, such as the coronary sinus and/or the His bundle region, to define the plane of the posterior atrioventricular groove and the anterior septum. The LAO fluoroscopic angle should be adjusted to have the His bundle, which marks the anterior septal plane, perpendicular to the plane of the screen; the RAO projection is then adjusted to be perpendicular to the LAO projection, with the coronary sinus catheter foreshortened. As mentioned, the His bundle catheter is an important reference to mark the site of the aortic root; alternatively, some investigators position a pigtail catheter in the aortic root as a fluoroscopic landmark. The pigtail catheter is usually positioned within the noncoronary sinus of Valsalva (the most posterior portion of the aorta).
When the transseptal approach is obtained via the standard route from the femoral vein, a long guidewire is typically inserted from the femoral vein into the SVC. The transseptal sheath-dilator assembly is then advanced over the guidewire until the dilator’s tip reaches the level of the tracheal carina (see Fig. 39.6 ). The guidewire is then withdrawn, the transseptal dilator is flushed with heparinized saline, and the transseptal needle (flushed with heparinized saline and connected to a small syringe with contrast) is inserted into the sheath-dilator system. As mentioned, the standard Brockenbrough needle has a pointer on the proximal hub, which indicates the direction of the needle; the transseptal needle should be advanced under fluoroscopy and positioned when the tip is just inside the distal end of the dilator. The proper orientation of the pointer hub (and needle tip) largely depends on the individual patient’s anatomy, although a position between 4 o’clock and 6 o’clock is typically adequate to maintain the needle perpendicular to the plane of the fossa ovalis. At this stage, the transseptal sheath/dilator and needle system are slowly withdrawn; when crossing the SVC-RA junction, the tip dives inferiorly and posteriorly (i.e., rightward in the LAO projection) (see Fig. 39.6 , 
). Once in the RA, its characteristic pressure waveform is recorded. At this point, before further withdrawing the system, the orientation of the dilator tip should be confirmed in the RAO and LAO projections. In RAO, the orientation of the dilator tip should be posterior to the His bundle (or pigtail catheter positioned within the noncoronary sinus of Valsalva) catheter and parallel or slightly posterior to the plane of the coronary sinus catheter. Once appropriately oriented in the RAO projection, the system is further withdrawn in the LAO projection. A second drop is typically seen when passing from the muscular IAS into the fossa ovalis (see Fig. 39.6 , see 
). Once the fossa ovalis is engaged, the direction of the dilator tip is reconfirmed in the RAO projection. Once the proper orientation of the dilator tip is achieved, the needle is advanced slowly into the fossa ovalis in the LAO projection. In the absence of pressure transduction, small quantities of dye may be injected through the transseptal needle ( Fig. 39.7 , 
). If correctly positioned, the contrast stains the fossa in a curtain-like fashion, highlighting the tenting of the fossa ovalis. If no tenting is observed or if contrast injection results in diffuse staining of the septum, the needle is likely at the level of the muscular septum. To avoid complications such as cardiac perforation or intramural hematoma, when appropriate positioning cannot be confirmed, it is advisable to repeat the whole maneuver to engage the true fossa ovalis. This requires withdrawing the needle from the body, reintroducing the guidewire into the sheath-dilator system, readvancing the system to the level of the tracheal carina, and repeating the process of engaging the fossa ovalis. Sometimes, when pushing the needle against the fossa ovalis, the whole system moves up toward the muscular rim instead of tenting the fossa ovalis; this typically indicates that the curve on the transseptal needle is insufficient and should be manually adjusted, by bending the mid-distal shaft of the needle to better reach the septum. Once the needle is appropriately tenting the fossa ovalis, the needle is advanced against the fossa ovalis and gently pushed forward until it enters the LA. An additional contrast injection confirms the correct positioning within the LA (the contrast dilutes within the LA and enters the LV) (see 
). Correct positioning of the needle in the LA may also be obtained by confirming an appropriate LA pressure tracing (see Fig. 39.7 ). If the injected contrast passes upward or a systemic arterial pressure tracing is recorded from the transseptal needle, puncture of the aortic root is likely. It is imperative that inadvertent aortic puncture is recognized to avoid introducing the transseptal dilator and/or sheath into the aorta. Once the transseptal needle is within the LA, counterclockwise torque may be applied to the whole system to prevent inadvertent puncture of the LA posterior wall (see Fig. 39.2 , 
). It is useful to then gently advance the entire system into the LA with the needle exposed until the dilator passes through the fossa ovalis. In this manner, the tented needle serves to stabilize the fossa and thereby facilitates the crossing of the larger diameter dilator. Once the dilator is positioned within the LA, the needle is withdrawn just inside the tip of the dilator. It is very important that the needle remains “covered” by the dilator to prevent inadvertent puncture of the LA wall; however, the needle should remain within the LA near the tip of the dilator (i.e., on the left atria side of the septum) to provide support for the subsequent crossing of the transseptal sheath. Failure to maintain adequate support from the dilator/needle combination within the LA will often result in failure to cross with the transseptal sheath and possibly the loss of LA access. When the transseptal sheath crosses into the LA, the TSC apparatus is seen to dive posteriorly (i.e., rightward in the LAO projection) toward the lateral LA wall; tactile feedback also typically accompanies sheath crossing. Once the sheath enters the LA, the operator should reflexively withdraw the dilator and needle into the sheath to minimize potential trauma to the atrial wall ( 
). If tip pressure monitoring is used, as long as a clear LA waveform is recorded, the tip is in the body of the LA (or in a PV) and not against a wall. The dilator and needle are then withdrawn together under continuous aspiration to prevent entrainment of air within the transseptal sheath and subsequent air embolism.
Intracardiac Echocardiography Guided Transseptal Catheterization
There are two commercially available ICE transducers: mechanical (i.e., radial) and phased array systems. The radial ICE transducer is mounted on a 9 F, nonsteerable catheter and emits an imaging beam at a 15-degree forward angle, perpendicular to the long axis of the catheter (Ultra ICE TM , Boston Scientific, San Jose, CA). The transducer rotates at 1800 rpm and produces a 360-degree imaging plane. The fixed, 9-MHz transducer frequency provides excellent near-field resolution; however, far-field structures are poorly visualized with radial ICE necessitating imaging proximate to the structure of interest ( Fig. 39.8 , 
).
