INTRODUCTION

Atrioesophageal fistula (AEF) is a well-described, rare, and serious complication of atrial fibrillation (AF) catheter ablation with associated high mortality rates (Figure 46.1). The incidence of confirmed AEF after AF catheter ablation is 0.015% to 0.04% based on surveys, but this figure is likely underreported due to a number of factors, including misdiagnosis and low response rate.^1–3 The AEF mortality rate has been reported to be 40% to 100%.^3–6 The high mortality rate related to this complication is likely in part due to lack of clinical awareness leading to delayed diagnosis, in addition to the complex nature of surgical repair required for treatment.

Figure 46.1 Pathologic specimen of atrioesophageal fistula (AEF) with esophageal (E) left inferior pulmonary vein (LIPV) fistula (arrow). Reproduced with permission from Khakpour et al. Atrioesophageal fistula after atrial fibrillation ablation: A single center series. J Atr Fibrillation. 2017;10:1654–1658.

Patients usually present 3 to 6 weeks after the index ablation with a constellation of infectious, neurologic, and thoracic findings including fever, leukocytosis, sepsis, neurologic deficits, and chest pain.5,6 Diagnosis can usually be made reliably with CT scan of chest with intravenous contrast; pneumomediastinum and air in the left atrium are the most common CT findings. This imaging modality has been shown to be the most useful tool in multiple prior studies.^7,8 Diagnostic esophagogastroduodenoscopy should not be performed prior to surgical repair due to the inherent risk of worsening air emboli from endoscopic air insufflation required during examination. There is limited data regarding the prognosis of those patients who survive to undergo surgical repair due to the small number of reported cases, but it appears to be favorable.^9,10 Based on the data available, the type of repair seems to affect prognosis, with those receiving esophageal stenting having the worst outcome and those undergoing surgery with combined left atrial and esophageal repair having the best result.^7,11

Similarly, esophageal injury during epicardial ablations has been described. This is due to the close proximity of the posterior left atrial wall and posterior mitral annulus to the esophagus. The relationship between the esophagus and the left atrial posterior wall and basal inferior left ventricular (LV) wall is variable, and the esophagus is most susceptible to injury where it is closest to areas of epicardial ablation. The esophagus is located in a groove posterior to the left atrium bounded by the thoracic vertebral column and the aorta posteriorly. Thermal injury during ablation is thought to affect the esophageal microvasculature, leading to ischemic necrosis and ulceration. Progression from ulceration to AEF formation may be facilitated by esophagitis, adjacent fatty necrosis, gastric hypomotility, and periesophageal vagal plexus injury, resulting in lower esophageal sphincter relaxation and acid reflux.^12–16 The time period for these subacute changes to occur could explain the delayed presentation of AEF after ablation.

Kik et al. reported 3 cases of AEF after minimally invasive video-assisted epicardial ablation for AF with bipolar radiofrequency (RF) energy.¹⁷ Another case of AEF was reported in a patient undergoing concomitant mitral valve replacement and a maze procedure using a unipolar RF source.¹⁸ Patients presented 3 to 6 weeks after surgery with either fever or neurologic symptoms. Esophageal perforation has also been reported during AF ablation using epicardial high-intensity focused ultrasound (HIFU).¹⁹ Esophageal injury during epicardial left ventricular tachycardia (VT) ablation has also been described. Koruth et al. reported a case of pericardioesophageal fistula after epicardial VT ablation with bipolar RF energy in the region of inferior mitral annulus.²⁰ Another case described a mediastinal-esophageal fistula after an epicardial VT ablation diagnosed 15 days after the procedure.²¹

These reports highlight the importance of awareness of esophageal injury when epicardial ablation in the posterior left atrial wall and basal left ventricle is pursued. Preventive measures to reduce risk of esophageal injury and, when indicated, interventions for esophageal protection should be utilized.²²

PREVENTION METHODS

Proton Pump Inhibitors

Experimental and clinical data suggest a role for gastroesophageal reflux in the pathogenesis of AEF. Esophageal injury, including esophageal ulceration, as documented by endoscopy, occurs at a high rate after AF ablation. Esophageal injury was reported in 15% of AF patients undergoing RF ablation in one study.^13,23 Gastroesophageal reflux may aggravate these injuries and in part lead to development of AEF. Gastroesophageal reflux may also develop or be exacerbated after ablation due to gastric hypomotility and periesophageal vagal plexus injury, resulting in lower esophageal sphincter relaxation. Experimentally, Yokoyama et al. demonstrated that esophageal ulcer progression and development of AEF in two dogs following ultrasound ablation were associated with reflux esophagitis.²⁴ One study demonstrated a positive correlation between symptomatic reflux and endoscopically diagnosed esophageal wall changes following radiofrequency ablation for AF.²⁵ Another study evaluated 31 patients undergoing RF ablation for AF using a leadless pH-metry capsule. A substantial number of patients (5/26, 19.2% respectively) without reflux prior to ablation acutely developed gastroesophageal reflux diagnosed by a pathologic DeMeester score.²⁶ The 2017 AF ablation consensus statement recommends use of proton pump inhibitors as a preventive strategy for AEF.²⁷

Esophageal Visualization

Preprocedural imaging modalities such as cardiac magnetic resonance imaging and computed tomography (CT) provide information about the course of esophagus in relation to the heart and the areas of ablation. Real-time visualization tools include fluoroscopy after administration of barium contrast in the esophagus or after placement of an esophageal temperature probe, as well as nonfluoroscopic localization using 2D and 3D echocardiography or using esophageal temperature probes with an integrated sensor for visualization with 3D mapping systems (Figure 46.2).^28–31

Thermal conduction during ablation is the primary cause of esophageal injury, and esophageal lesion is mainly influenced by the proximity of the energy source and the thickness of connective tissue.32 This in turn has led to the rationale for using luminal esophageal temperature (LET) monitoring as a surrogate for injury during ablation in its vicinity.³³ Varying data exists regarding correlation between maximal detected temperature and development of injury. Cummings et al. showed that during left atrial RF ablation using a single-sensor esophageal probe, esophageal temperature was significantly higher during lesion application over the course of the esophagus, and lesions that generated early changes of esophageal injury as evaluated by endoscopy had higher esophagus temperatures than those that did not (39.3 ± 1.5° C, 38.5 ± 0.9° C, P < 0.01).²⁸ Singh et al. showed an 83% relative risk reduction in development of esophageal ulceration with LET monitoring when ablation was interrupted at LET of 38.5° C.³⁴

Despite the use of temperature monitoring, esophageal injury can still occur.³⁵ In fact, some studies have suggested that routine LET monitoring may contribute to esophageal damage due to thermal effects,^36,37 though this was not confirmed in a computational modeling study in which no thermal interaction was found between the esophageal temperature probe and the ablation catheter.³⁸ Nonetheless, LET monitoring limitations include difficulty in maintaining a close proximity of the probe to the catheter and unpredictable distance of the temperature probe from esophageal adventitia tissue, which is the site of highest esophageal temperature. Multiple different singlesensor and linear and sinusoidal multisensor esophageal temperature probes are currently available. Multisensor probes have been employed in an attempt to decrease the variable distance between catheter and probe and some studies have shown a higher rate of elevated temperature detection with use of the multisensor probes as compared to single-sensor ones.³⁹ The 2017 AF ablation consensus statement recommends the use an esophageal temperature probe during RF ablation procedures to monitor esophageal temperature and help guide energy delivery (class IIA, LOE-C).²⁷ It is currently the authors’ practice to stop ablation when a rise of 1° C in LET from baseline is noted, or if a temperature of 39° C during ablation is reached.

Mechanical Displacement of Heart Using Epicardial Balloon

Stay updated, free articles. Join our Telegram channel

Join