Esophageal Injury during RF Ablation Procedures

Abstract & Commentary

By John P. DiMarco, MD, PhD

Source: Singh SM, et al. Esophageal injury and temperature monitoring during atrial fibrillation ablation. Circulation: Arrhythmia Electrophysiology. 2008;1:162-168.

In this report, Singh et al report a two hospital study on the use of esophageal temperature monitoring during atrial fibrillation ablation procedures. This was a nonrandomized study that included 81 patients who underwent AF ablation followed by a post procedural, not symptom-driven esophageal endoscopy within three days of the ablation procedure. Sixty-seven patients underwent luminal esophageal temperature (LET) monitoring during their ablation procedure, whereas 14 did not. LET monitoring was performed using a 9Ff single thermocouple esophageal temperature probe that was inserted and advanced into the esophagus using fluoroscopic guidance. The position of the tip of the temperature probe was monitored fluoroscopically throughout the procedure and was adjusted to the level of the ablation catheter before application of each ablation lesion. Procedures could be performed either under conscious sedation or general anesthesia. Ablation lesions were placed with either internally or externally irrigated catheters. RF energy delivery was set to a maximum of 35 W and 400 C. RF applications were terminated if the LET exceeded 38.50 C in patients. Esophageal ulceration was attributed to the AF ablation if it was located on the anterior wall of the mid esophagus adjacent to the pulsating heart. Patients with ulcerations were treated with high-dose proton pump inhibitor therapy for one week with repeat endoscopy to ensure ulcer healing.

The study population was an average group undergoing an ablation for atrial fibrillation. There were 81 patients with a mean age of 58 years. Paroxysmal atrial fibrillation was present in 44%, with a mean duration of atrial fibrillation of over four years. Sixty-five percent had a structurally normal heart. The mean left atrial size was 43 mm. Procedures were performed under general anesthesia in 13% of the patients with LET monitoring vs 43% of those without LET monitoring. Although radio frequency energy application was stopped, if the LET reached 38.50 centigrade, higher temperatures were noted in half of the patients. There was not a clear relationship between the maximum LET recorded and the presence of ulcerations. However, a trend toward more ulceration was noted in patients with maximum LET ≥ 39°. In the LET monitoring group, four of 67 patients (6%) had esophageal ulcers visualized during the follow-up endoscopy. In contrast, five of 14 patients without LET monitoring were found to have post-procedure esophageal ulcers. An interesting but unexplained finding was the presence of pericarditis or pericardial effusion in six of 67 patients with LET monitoring vs none of the patients without LET monitoring. The AF recurrence rate was 31% among the patients who underwent LET monitoring and 43% in those without LET monitoring.

Ulcerations in patients without LET monitoring appeared longer and more linear compared with ulcerations observed in patients who had LET monitoring. All ulcerations had healed by the follow-up EGD at one week. Left atrial esophageal fistulae did not develop in any patient.

Singh et al conclude that LET monitoring is straightforward, inexpensive, well tolerated, and may decrease the risk of esophageal thermal injury during catheter ablations for atrial fibrillation.


Atrio-esophageal fistula formation is a rare but life-threatening complication of ablation procedures for atrial fibrillation. Patients who develop fistulae present with sepsis and air embolism usually one to two weeks after the procedure; the mortality rate is very high. The mechanism for the fistula formation is thought to be thermal injury to the esophageal muscular wall during the procedure. It can be seen with all types of ablative energy, including radio frequency, high-intensity-focused ultrasound, and cryothermy. The esophageal wall lies in close proximity to the posterior wall of the left atrium. Since the left atrial wall is quite thin, thermal injury to the muscular wall of the esophagus may be difficult to avoid when posterior left atrial structures need to be ablated.

There are a number of different approaches that may be used to prevent esophageal injury. Various imaging techniques, including fluoroscopic visualization of barium in the esophagus, intracardiac ultrasound image monitoring, and correlation with pre-procedure CMR and CT scans may be used. The operator can reduce the energy when the esophagus is in close proximity to the ablation catheter. Increased mobility of the esophagus, seen in patients who undergo ablation using conscious sedation, compared to when general anesthesia is used, may also be protective. The technique described here of luminal esophageal temperature monitoring may have some value but also some significant practical limitations. It may be difficult to position the single thermistor precisely at the point of maximum temperature during lesion placement, and one could be fooled if folds of the esophagus or differences in projection caused one to miss the area of peak injury. The major observation from this paper is the relatively high rate of esophageal ulceration noted. Overall, in the entire series, 11% of the patients developed esophageal ulcerations. Fortunately, these were all asymptomatic and resolved quickly with proton pump inhibitor therapy. This high frequency of injury, however, should make electrophysiologists performing catheter ablations even more carefully to guard against the possibility of esophageal damage.