Improved Anatomical Orientation During AF Catheter Ablation: Experience from Leipzig Heart Center

As one of the premier heart centers in Germany, our institution performs catheter ablation with great frequency. Approximately 2,000 ablation cases per year are going to be performed in two EP labs, with an increasing number of those cases being AF patients. Three-dimensional mapping, atrial chamber reconstruction and non-fluoroscopic catheter visualization are being used in most of these cases. In May 2007, we began using the EnSite System (St. Jude Medical, St. Paul, Minnesota) with the EnSite Fusion Registration Module. To date, our group has completed nearly 100 cases using this software. Although we have only recently become regular users of this system, the introduction of EnSite Fusion has had a very positive impact, and we now find it very easy to use. It provides us with a realistic understanding of individual patient anatomy, and facilitates the ablation procedure by enabling us to work entirely in the registered CT model without any additional chamber surface reconstruction. This accurate anatomical orientation is difficult to achieve with electroanatomical chamber reconstruction alone, where the resolution of the anatomical model is limited in regions with difficult catheter access. With the EnSite Fusion tool, we are able to achieve an accurate, quality registration in no more than 10 minutes, a fact which has definitely helped reduce our average procedure time. Registration of the CT Image into the Mapping System CT scans are performed on a 64-row spiral CT scanner and segmented prior to the procedure. For re-ablation cases, we use the initial CT scan from the index procedure. Segmentation is performed using the Verismo tool on the EnSite System. We begin all AF ablation procedures by placing patients in deep sedation, then inserting standard catheters in the coronary sinus (CS) and right ventricular apex (RVA). We then perform a single transseptal puncture using the Agilis NxT Steerable Sheath (St. Jude Medical). Once access is gained into the left atrium (LA), we begin IV heparin to achieve ACT of 250 ms. The sheath is continuously flushed. Initially, a 20-pole Optima spiral mapping catheter is inserted into the left atrium and optimized. The spiral catheter is used to access each pulmonary vein, using the synced CT image as anatomical guidance. Each vein is assessed for conduction, then the catheter is dragged out of the PV as points are taken through all 20 electrodes to create individual PV geometries. This is done for all veins, and afterwards the spiral catheter is removed and saved for later use (Figure 2). At this point the ablation catheter is inserted. We utilize the Coolpath (M-type, 4 mm irrigated tip) ablation catheter (St. Jude Medical). The catheter is inserted into the left lower PV and then slowly dragged out until it falls into the left atrial cavity. The inferior edge of the left lower PV is taken as a landmark (fiducial or anchoring point) on the CT and real-time models. Using this point, we then complete the initial registration of the CT anatomy into the real-time model (Figure 3). After this initial step, we improve the registration by further aligning the CT PVs with the reconstructed PV anatomies. Beginning with the vein appearing furthest from the CT surface, we add in on average four registration points at corresponding locations between reconstructed PV anatomy and CT surfaces (Figure 4). Our objective is that all reconstructed PVs appear within the CT image (Figure 5). It is important to keep in mind that points should be added so that the CT is brought towards the reconstructed PV anatomies at this phase. The software allows adjustment of the registration in either direction. When this objective is achieved, we hide the reconstructed PVs and remove unnecessary distal parts of the CT PVs to maintain image clarity. We then further improve our registration by placing the ablation catheter at the LA roof and using it to confirm proximity to the CT model (Figures 6A-C). If we see a location with inexact alignment (e.g., catheter is not at the CT surface), we add a landmark point pulling the CT image onto the catheter tip (this does not alter the CT model.) This is repeated at the basal posterior wall and the LA isthmus. Our technique for CT image registration is based on the concept that the pulmonary veins are the most stable anatomical structures in the heart. Therefore, we have based our major fixation for the CT image on the pulmonary vein location, and then fine-tuned the registration process with a few additional left atrial landmark points. Using the Verismo software allows for an accurate three-dimensional reconstruction of the CT image for a wide range of quality of the initial CT data. Even in single cases scanned without usage of contrast medium, we were still able to reconstruct the left atrium. It is important to remember that while the anatomical accuracy and detail of a CT image are valuable, the ability to make fine adjustments is critical for effective catheter navigation. Registration is not always 100% accurate. Dynamic Registration allows us to adjust the registration and accommodate the natural variations that occur between the time of CT scanning and the procedure. Application of Therapy and Confirmation of Endpoint At the Heart Center, we primarily take an anatomical approach to ablation. Patients presenting with paroxysmal AF receive circumferential left atrial ablation lines, and those with persistent AF receive an additional box lesion and a mitral isthmus line (Figure 7). Patients presenting with AF at the beginning of the procedure receive an electrical cardioversion, and ablation is performed during sinus rhythm (SR). Prior to ablation, we insert a catheter into the esophagus. We utilize an Esotherm catheter with three temperature probes (Fiab, Florence, Italy), which allows us to visualize the esophageal location (through EnSite geometry reconstruction or direct catheter visualization) and to simultaneously record esophageal temperature along the posterior left atrial wall (Figure 8). We measure the esophageal temperature throughout the case, and have found it to be very sensitive. Our practice is to reduce energy output to 25 watts when ablating in close vicinity to the esophagus, and we are quite cautious with continuous monitoring of the esophageal temperature in these areas. In fact, we have seen remarkable temperature rises even with 20 to 25 watts, which necessitated further energy reduction. As we create circumferential lesions around the left and right PVs, the ablation sites are tagged with 3D lesion markers, which represent the true position of the catheter tip in three-dimensional space (no projection to surface or lesion at mouse) and confirm registration accuracy in case of true surface contact to the CT image. In areas with minor discrepancies between catheter location and CT surface, we can quickly adjust the CT registration during the procedure. The position of nearly all 3D lesion markers on the CT surface at the end of the procedure reflects the high degree of accuracy of the registration process and the subsequent accurate anatomical orientation (Figures 9A-E). We utilize the Agilis steerable sheath for most movement within the atrium; the ablation catheter is only handled passively (advanced or retracted for optimal wall contact). We have found this to enhance stability of the catheter when creating lesions, and to enable improved catheter access to challenging areas of the left atrium. For instance, maintaining good position on the ridge between the left upper pulmonary vein and the left atrial appendage can be quite difficult, but we have been able to place the catheter there with good stability and reproducibility using the steerable sheath. While the patient is in sinus rhythm, we place circumferential left atrial ablation lines around the left- and right-sided PVs. Following isolation of the whole PV antrum, we confirm line continuity by applying maximum output stimulation through the ablation catheter while mapping along the inner aspect of the circumferential ablation line and checking to see if this captures the LA (seen on the CS signal). If the LA is captured, this technique also clearly shows where the gap in the line exists, whether it is near the superior, posterior, inferior or anterior sections of the lesion. When gaps are found, we re-ablate until we lose capture. One of the benefits of this approach is that we are able to complete the ablation process using a single transseptal puncture, a single sheath, and a single catheter. We usually reconfirm bi-directional conduction block in and out of the PV antrum using the spiral catheter; however, the results are always in concordance with the ablation catheter's PV antrum mapping. After isolating the left- and right-sided PV antrum in patients with persistent AF, we place a LA roof line and a line at the basal posterior LA (connecting the LLPV and RLPV). This way, a box at the posterior LA wall is created and complete isolation of the whole box is again achieved and verified using voltage and pace mapping. Eventually a left atrial isthmus line is placed as continuously and transmurally as possible. Line continuity is not enforced with additional CS ablations. At the end of the procedure, we test for AF inducibility with 20 seconds of aggressive burst stimulation at the atrial refractory period; this is currently still without consequences with respect to additional lesion creation. Our procedural endpoint is the electrophysiologically-defined complete electrical isolation of all PV antrums, and in patients with persistent/ permanent AF, additional isolation of the posterior box lesion together with a mitral isthmus line. Key Points and Conclusion One must acknowledge that there is a learning process with any new technology. As we were not frequent users of the EnSite System prior to the availability of Fusion, we had to become familiar with the interface, but this was quickly accomplished and we now find the system easy to use. It is important to take advantage of clinical support when learning the system, and to focus on the information that is important for the procedure. There is a great deal of information available through this mapping technology, but it is not necessary to display everything on the screen. Most importantly, stay open-minded and develop the techniques that work best for you and your patients. To help us get accustomed to the EnSite system, we began by using the system on simple arrhythmias such as atrial flutter or AVNRT. Now AF is our major indication for use of the 3D system, and we are also using it regularly in more complex arrhythmias such as atrial macroreentrant tachycardias. In those cases we needed to introduce information derived from electrophysiological assessments such as activation, entrainment and voltage mapping, all of which can be nicely displayed on the integrated CT image, also without 3D chamber reconstruction. This provides a very fast and accurate understanding of the true anatomical location of the reentrant circuit. The basic ablation concepts described in this article have been used in our institution for some time. With the introduction of new technological capabilities, we are simply able to apply them better in the individual patient with his or her individual left atrial anatomy. Finally, because we are able to achieve this registration in a very short amount of time, it appears that this technology may provide the important benefit of saving time in our very busy EP lab.