Strategies to Minimize the Risk of Esophageal Injury During Catheter Ablation for Atrial Fibrillation

Tristram D. Bahnson, M.D.


Pacing Clin Electrophysiol. 2009;32(2):248-260. 

In This Article

Catheter Ablation for AF Using a Combination of RF and Cryothermy Ablation -- a Practical Approach

As previously discussed, an operator has limited ability to define all variables that determine RF-induced heat transfer to the esophagus, and there are no proven methods to monitor for heat transfer to and thermal injury of the esophagus during ablation. Accordingly, specific assessment of the potential for or occurrence of esophageal injury during RF ablation in an individual cannot reasonably be made in clinical practice, and it may be safest to assume that heat transfer to the esophagus will be efficient and the risk of esophageal injury high whenever delivering RF energy near the esophagus. Given these uncertainties the author uses the following approach to minimize risk of esophageal injury during catheter ablation:

We assume that the potential for efficient heat transfer to and thermal injury of the esophagus is high whenever any of the following are noted: (1) the MRI (or CT) shows the esophagus to overlay the right or left PV antrum and/or ostium, (2) rapid mural esophageal heating is suggested by recording an early and rapid rise in luminal esophageal temperature during an RF pulse, or (3) real time monitoring with ICE or fluoroscopic visualization of the luminal probe suggests that the intended site of ablation is within 1-2 cm of the esophagus. When ablation targets are so deemed risky, our approach is to employ the above strategies alone or in combination.

Preprocedure imaging with CT or MRI is done in all patients to provide baseline imaging of the PV anatomy and morphology. These studies often allow for an initial assessment of the location and breadth of the esophagus-atrium contact region. In particular, the finding of a broad esophagus covering much of the posterior left atrial wall recommends caution when using a discrete luminal marker like a temperature probe as the sole modality to define the location of the esophagus during ablation. During the procedure the esophagus location relative to the atrium is displayed using fluoroscopy of a luminal marker, a segmented shell of the esophagus from CT when available as a registered structure (Figure 2, panels F and G), and ICE to confirm location of the esophagus in real time. Recent implementation of Carto-Sound™ is helpful to achieve informative real-time imaging during ablation. As already considered above, a temperature probe or catheter might well be positioned eccentrically within the esophagus lumen, suggesting that any endocardial ablation sites within 1-2 cm medially or laterally to the fluoroscopically defined luminal probe could in fact be immediately adjacent to esophageal tissue (Figure 3) unless real-time ICE imaging clearly indicates that a particular site visualized to be near the luminal marker with fluoroscopy is not. ICE imaging from the right atrium has some limitations in making this determination due to the narrow 2-D imaging plane and the likelihood that the catheter tip and adjacent esophagus tissue, each viewed obliquely from the right atrium, might not fall within a single imaging plane even when the catheter tip is adjacent to some part of the esophagus.

When a potential ablation target is deemed risky, we typically attempt limited ablation near the esophagus with one or two RF pulses limited to less than 20-25 W, and lesion duration limited to less than 20 seconds, independent of readings from the luminal temperature probe tip that is actively positioned immediately adjacent to the ablation site. If a rapid temperature rise is recorded by the luminal probe, suggesting an instance of efficient heat transfer to the esophagus, RF energy delivery is immediately discontinued and subsequent RF energy delivery at this site avoided. Lack of a rapid temperature rise is not used as justification for more extensive RF energy delivery at ablation sites deemed near to the esophagus by other criteria. If additional ablation is required to achieve PV isolation or to ablate tissue at the posterior PV antrum thought to be arrhythmogenic, the author replaces either the lasso catheter or RF ablation catheter with a 6-mm-tip cryothermy ablation catheter (CryoCath, Inc). In the former case, the retained mapping catheter is used to tag cryothermy ablation sites on the electroanatomic map as they are delivered (Figure 5, panels A-C), and in the latter case a fluoroscopic image of each cryothermal site relative to the circular mapping catheter is saved (Figure 5, panel D), and the sites later marked under fluoroscopic guidance after the cryoablation catheter are again replaced by the electroanatomic mapping catheter (Figure 2, panel B). Other investigators (Ken Ellenbogen, M.D., pers. comm.) place a 10.5 French sheath in the beginning of the case if preoperative anatomy suggests the esophagus is located close to the pulmonary veins, or uses a stiff guidewire to exchange the 8 or 8.5 French transseptal sheath with the larger sheath to allow introduction of a large-caliber 8-mm-tip cryothermy catheter. This catheter is capable of greater refrigerant flow and would be expected to produce larger cryothermy lesions per unit time of application. The cryothermy catheter location can alternatively be monitored directly when the NavX™ (St. Jude Medical, Fullerton, CA, USA) system is used. Target cryothermy ablation parameters employed by the author include tip temperature ≈-75°C for 250 seconds with one or two cycles of cryothermy at each ablation site. This approach results in a limited segment at the posterior wall being targeted with cryothermy and all remaining ablation accomplished with RF energy, and, in our experience, has resulted in longstanding PV isolation. Significant cooling of the esophagus, as previously reported,[17] is often observed during cryothermy ablation consistent with the targeted site being a risky one for RF ablation had extensive RF energy been employed, due to thermal coupling with the esophagus. Figure 5 shows two representative patients in whom the esophagus was located near intended ablation targets and in whom a hybrid approach was used employing RF ablation for all regions except those adjacent to the esophagus that were targeted with cryothermal energy. When important sites of ablation are identified to be near the esophagus by any imaging methods, or by observing rapid elevation of luminal esophageal temperature during initial RF energy delivery, and isolation of the PV and PV antrum cannot be achieved with limited RF energy delivery (e.g., 1-2 low power and brief RF energy pulses), it may be reasonable to consider an alternate ablation strategy such as cryothermy. Prospective outcomes studies will be required in larger numbers of patients to define the relative risks of specific ablation strategies to result in clinically significant esophageal injury.

Figure 5.

Examples of hybrid therapy employing cryothermy for ablation when adjacent to the esophagus. Panels A through C are from the patient depicted in Figure 1, panel D. In the LAO and RAO views, respectively, the luminal probe is marked with white arrows, and the location of the cryothermy catheter is confirmed by briefly positioning the Carto catheter (Thermocool™) at the cryothermy site. Cryothermy sites are tagged in blue as seen in panel C on the E-A map with a superimposed segmented and registered shell of the left atrium from a previous MRI image. The MRI image does not include the entire left atrium in this example. Panel D shows cryothermy ablation at the right inferior PV guided by a circular mapping catheter positioned at the vein os in the patient from Figure 2, panels A-D. Sites of cryothermy were tagged after lesion delivery as seen in Figure 2, panels B-D based on biplane fluoroscopy with the circular mapping catheter used as reference. White arrows point to the luminal temperature probe. Catheters are also seen in the coronary sinus, and RV (panel D only). The ICE catheter is in the RA. Panels E and F are from the patient depicted in Figure 2, panels E-G. These 2-D ICE images were obtained in real time during cryothermy at the left superior PV antrum in proximity to the esophagus, and represent two frames of an acquired image loop less than one second apart. The left atrium (LA), aorta (Ao), and posterior margin of the esophagus (thick arrows) compressed between the LA and Ao, are marked. The cryothermy catheter tip is seen (thin arrows) contacting the posterior atrial wall. The high frequency of the refrigerant pump used by this cryothermy system (CryoCath, Inc.) causes a blurring artifact of the catheter. An echogenic spot at the catheter tip in panel E depicts the forming "ice ball," and adherence of the catheter to the endocardium is depicted in panel F by the "tenting" of the tissue at the catheter-tissue interface due to motion of the heart relative to the catheter during the cardiac cycle.