Active esophageal cooling reduces peri-esophageal and vagal nerve injury in radiofrequency-current pulmonary vein isolation

Abstract

Background

Damage to peri-esophageal tissue may occur following pulmonary vein isolation (PVI). Active esophageal cooling has been shown to reduce the incidence of mucosal esophageal injury, probably by dissipation of heat and inhibition of inflammation. Whether it also protects the peri-esophageal vagal nerve plexus and reduces gastric hypomotility and food retention is uncertain.

Objective

The study aimed to analyze and compare the incidence of esophageal and vagal nerve injury following radiofrequency-current (RF) PVI with active esophageal cooling to that of luminal esophageal temperature (LET) monitoring.

Methods

Using endoscopy and electrogastrography, esophageal and peri-esophageal injury (mucosal lesions, food retention, and vagal nerve injury) were prospectively assessed following RF-PVI with active esophageal cooling and compared with RF-PVI with LET monitoring.

Results

A total of 64 patients (69 [65/75] years, 58% men) undergoing RF-PVI for atrial fibrillation with esophageal cooling under (deep) conscious sedation were prospectively studied and compared with 52 LET-monitored patients.

Following RF-PVI with active cooling, 4.7% had mucosal erythema, 15.6% new-onset food retention, and 14.1% ablation-induced vagal nerve injury. In comparison, the LET-monitored cohort showed 11.5% with mucosal esophageal lesions, 26.9% new-onset food retention, and 28.8% ablation-induced vagal nerve injury. The rate of any esophageal injury per patient was decreased by a factor of 0.51 ([95% confidence interval 0.30, 0.86]; P = .0142).

Conclusion

In RF-PVI, active esophageal cooling reduces ablation-induced vagal nerve injury and overall peri-esophageal injury.

Graphical abstract

Key Findings.

• ▪Active esophageal cooling in radiofrequency pulmonary vein isolation (PVI) reduced the incidence of mucosal esophageal lesions, vagal nerve injury, and new-onset food retention by about 50%.
• ▪Similar to PVI without cooling, post-procedural vagal nerve injury incompletely overlapped with other manifestations of peri-esophageal injury.
• ▪Active esophageal cooling allowed shorter procedure times and was used successfully in patients undergoing PVI under sedation with unassisted respiration.

Introduction

Pulmonary vein isolation (PVI) is a valuable treatment for patients with atrial fibrillation (AF). However, a serious drawback of thermal AF ablation is the risk of esophageal lesions (ELs) and atrio-esophageal fistula (AEF)¹ resulting from the close proximity of the left atrial posterior wall (LAPW) to the anterior wall of the esophagus.

The pathophysiology of EL progression to fistula is not well understood.² An initial thermal injury is likely a prerequisite for AEF formation, but a subsequent inflammatory cascade may also be pathophysiologically relevant for lesion progression to AEF.^3,4

Ablation-induced vagal nerve injury may lead to reflux⁵ and gastric motility disorders/food retention and enhance esophageal inflammation.⁶ Neuropathic alterations of gastric motility (as a consequence of vagal nerve injury) can be assessed by non-invasive registration of gastric electrical activity (electrogastrography [EGG]).^6,7 A previous study showed a 28% incidence of vagal nerve injury following radiofrequency-current (RF) ablation of AF.⁸

Preventive strategies (eg, luminal esophageal temperature [LET] monitoring, reduction of ablation power, and high-power short-duration concepts^9,10) so far have not eliminated the risk of fistula formation.^1,11

Attempts to prevent AEF by cooling the esophagus have been discussed as early as 2005.¹² Several studies have shown a significant reduction of thermal esophageal injury by esophageal cooling¹³ but it is not clear whether esophageal cooling protects the peri-esophageal tissue and vagal nerve plexus as well.

Additionally, studies to date on esophageal cooling have used general anesthesia, leaving uncertainty in utilization in patients undergoing these procedures under sedation.

Purpose

The study aimed to assess the incidence of esophageal and peri-esophageal vagal nerve injury following RF- PVI with active esophageal cooling compared with temperature monitoring.

Methods

The study was approved by the ethics committee of the Medical Association of Brandenburg/Germany [2323-65-BO-ff]. The research reported in this paper adhered to the guidelines of Good Clinical Practice principles and the Declaration of Helsinki. All patients provided written informed consent for participation in the study.

Consecutive patients undergoing RF-PVI between August 2023 and July 2024 at Medical University Lausitz – Carl Thiem, Cottbus, Germany, were enrolled. PVI was performed with an RF catheter with contact force measurement (Tacticath SE, Abbott, Chicago, IL). All interventions were performed under (deep) conscious sedation with fentanyl, midazolam, and propofol, inducing a Modified Observer’s Assessment of Alertness and Sedation (MOAAS) scale¹⁴ of 2 (rarely 1) with unassisted spontaneous respiration. Wide-antral-circumferential isolation of the ipsilateral pulmonary veins and additional left atrial (LA) lesions (at the operatoŕs discretion) were performed. Ablation power at the LAPW was set to 30 Watts. The target lesion size index was 4.5 at the LAPW and 5.5 in other areas.

Study cohort with active esophageal cooling

Following initiation of sedation and transseptal access (but prior to 3-dimensional electroanatomic mapping), the esophageal cooling probe (ensoETM, Haemonetics, Boston, MA) was inserted. The ensoETM device is a triple-lumen silicone tube of 73 cm in length and 1.2 cm outer diameter. Connected to a console (Blanketrol III, Gentherm Medical, Cincinnati, OH), distilled water circulates (2.4 Liter/min) in a closed loop using 2 of the lumens at a temperature of 4 °C. The water volume in the probe is 55 mL, exerting a maximum pressure of 103 kPa.¹⁵

Adequate depth of device placement covering the entire esophageal contact area with the LAPW is mandatory for its protective effect. (Figure 1) Cooling was initiated at least 10 minutes prior to the first RF lesion and continued until the end of the procedure. To monitor for systemic hypothermia, body temperature was recorded by a commercial in-ear thermometer. Forced-air warming was not used.

Figure 1.

Fluoroscopic image with the ensoETM probe. Fluoroscopic image in left anterior oblique (LAO 30°) angulation. The ensoETM probe is advanced into the stomach (radiopaque tip). The guidewire in the inner lumen of the cooling probe improves visibility and is removed after placement. The dotted line circumscribes the ensoETM probe (not visible on fluoroscopy). Catheters were placed at the His-bundle and in the coronary sinus, transseptal sheaths (TSP), and mapping catheter in the left superior pulmonary vein (HDgrid). The aortic valve had been replaced.

Control cohort

The control population was prospectively collected at the same center with results published as the Side Effects of PVI in AF (SEPIA) study.⁸ In those patients, LET was monitored by a shielded double-S-shaped probe with 12 thermocouples (S-Cath, Circa Scientific, LLC, Englewood, CO), and energy delivery was stopped when LET reached 41 °C. All other aspects of the ablation procedure, data collection, and diagnostic evaluation were identical.

Esophageal and peri-esophageal injury

For this study, esophageal and peri-esophageal injury included both endoscopically detected pathology and neuropathic patterns in EGG recordings. All patients underwent endoscopic evaluation (Olympus, GIF-Q165, Hamburg, Germany) of the upper gastrointestinal tract within 1 week before and within 2 working days after the PVI procedure. The pre-procedural esophagogastroscopy assessed pre-existing pathology and gastric food retention at baseline. Esophagitis, Barrett-mucosa, and esophageal erosion were considered signs of increased esophageal vulnerability (for ablation-induced ELs).¹⁶ In addition, pre-procedural endoscopy excluded anatomical obstacles for the placement of the cooling probe. The post-interventional study focused on new ELs facing the LA and new-onset food retention (after a fasting period of a minimum of 12 hours). Endoscopy was performed within 24 to 72 hours following PVI. Endoscopy within 24 hours might miss maturing mucosal lesions, whereas a delay ≥1 week underrates esophageal injury because of the resolution of mild lesions.¹⁷

EGG was performed to assess ablation-induced vagal nerve injury (neuropathic pattern). The technique of electrocardiography (ECG) is described elsewhere.⁸ Briefly, gastric electrical activity was recorded (MP160WSW-recorder and analog/digital converter, AcqKnowledge Software [Version 5.0.5], Biopac Systems, Goleta, CA) within 2 working days before and after PVI in the morning (and separated from endoscopy by more than 18 hours). Following a standardized meal, EGG was recorded in a food-activated state. The analyses, transformation, and calculations of the raw data were performed with Matlab R2019a (Mathworks, Natick, MA) and EGG DWPack Software.¹⁸

The stomach displays so-called “slow electric waves” triggering peristaltic contractions for the aboral food movement. Normally, the ingestion of food stabilizes the slow electric wave rate and increases the power of the dominant frequency.¹⁹

Ablation-induced thermic impairment of the peri-esophageal vagal plexus (the focus of this study) is associated with neuropathic EGG-pattern, defined as (1) new (or increased) gastric arrhythmia (absence of regularized slow electric wave rhythm) in the post-prandial period and/or (2) power amplification of slow electric wave ≤1% following food ingestion (compared with the power of the dominant slow electric wave frequency in the pre-prandial period).¹⁹ In addition, a combined analysis of mucosal lesions, food retention, and vagal nerve injury was performed to show the overall incidence of ablation-induced esophageal injury.

Statistics

The control cohort⁸ consisted of 52 patients with RF-PVI undergoing the same ablation protocol and diagnostic testing, but with LET monitoring in place of active esophageal cooling. This cohort showed a 28% incidence of ablation-induced vagal nerve injury. In the absence of preexisting data, we assumed a reduction by active esophageal cooling to a third (9%). With those assumptions, an alpha-error of 0.05 and a power of 0.8, a sample size of 64 patients was calculated to detect a significant difference in the rate of gastric arrhythmia and/or lack of post-prandial power increase (as signs of ablation-induced vagal nerve injury) compared with the EGG before ablation.

Continuous data were checked for normal distribution using Shapiro-Wilk test and are shown as median/interquartile ranges (1st/3rd quartile). Differences between groups were compared using Mann-Whitney U-test. All categorical data were assessed by Fisher exact test and are expressed as numbers (percentage).

In addition, the rate of the different patterns of esophageal injury (ELs, food retention, and/or vagal nerve injury) was analyzed. The analyzed rate per patient resulted in 0 (no injury at all), 1, 2, or 3 (all 3 patterns present in 1 patient). To compare the results, a Poisson regression with an overdispersion parameter (quasi-Poisson regression) was performed which provides a method robust to deviations from the Poisson variance for the counts. For this, the overdispersion parameter was estimated as close to 1 (0.98).²⁰ Use of the ensoETM cooling device was used as a covariate.

The analyses were carried out using SPSS-Statistics for Windows Version 29.0.0 (Armonk, NY, IBM Corp. USA), and R-statistics V4.4.1 (R-foundation, Free Software Foundation’s GNU General Public License). A 2-sided P-value < .05 was considered significant.

Results

A total of 68 patients were screened (hereof 13 with re-do interventions), and 4 had to be excluded (1 pericardial tamponade, 2 without LAPW ablation lesions [other than PVI], 1 missing EGG because of technical problems), providing a study cohort of 64 patients. Baseline characteristics of the study cohort and the control cohort⁸ are given in Table 1. PVI and LA linear lesions were completed successfully in all patients, and first pass isolation rate with cooling was close to 70% (similar to that of the LET-monitored cohort). Procedure duration, fluoroscopy duration, and LA dwell time were significantly shorter with active esophageal cooling, and other procedural, and ablation energy parameters (including total energy applied [∑ Energy] at the LAPW) were not different between cohorts. (Table 2)

Table 1.

Patient characteristics

Patient	Study cohort (n = 64)	Control cohort (n = 52)8	P value
Age (years)	70.8 (64.5/74.8)	69.1 (60.2/76.2)	.397
Men	57.8%	53.8%	.710
BMI (kg/m2)	29.6 (26.5/33.6)	28.0 (25.3/31.7)	.115
Atrial fibrillation
- paroxysmal	17.2%	28.8%	.180
- persistent	82.8%	71.2%
LA metrics
- LA Volume (mL)	131.0 (113.0/159.8)	136.0 (115.8/159.3)	.847
- LAVI (mL/m2)	65.7 (53.2/82.1)	69.0 (57.6/81.7)	.635
LV ejection fraction (%)	55.0 (53.0/60.0)	58.0 (53.0/65.0)	.115
Comorbidities
- Diabetes mellitus	21.9%	15.4%	.477
- Hypertension	85.9%	90.4%	.572
Ablation procedure
- PVI only	40.6%	46.2%	.577
- PVI + LA lesions	59.4%	53.8%

Baseline characteristics of the study cohort and the control cohort. There were no statistically significant differences between groups.

BMI = body mass index; LA = left atrial; LAVI = left atrial volume index in 3D-reconstructed computed tomography; LV = left ventricular; PVI = pulmonary vein isolation.

Table 2.

Peri-esophageal injury in the study cohort and the control cohort

Procedural data	Study cohort: (n = 64)			Control cohort (n = 52)8			p (2) w/wo cooling
	No injury	Any injury	p(1) w/wo injury	No injury	Any injury	p (1) w/wo injury
All (n)	45	19	P value	28	24	P value	P value
Procedure duration∗ (min)	150.0 (130.0/180.0)	150.0 (142.5/162.5)	1.000	165.0 (153.8/190.0)	160.0 (147.5/195.0)	.526	.016
LA dwell time (min)	120.0 (105.0/150.0)	125.0 (120.0/145.0)	.740	130.0 (118.8/166.3)	137.5 (110.0/161.3)	.971	.058
Fluoroscopy (min)	18.2 (15.2/22.1)	16.1 (13.3/23.1)	.252	21.8 (18.0/26.5)	24.4 (19.1/33.3)	.383	.002
RF at LAPW Duration (s)	1195.2 (908.0/1563.0)	1221.7 (888.5/1343.5)	.820	1313.5 (943.0/1544.5)	1087.2 (754.5/1338.8)	.259	.532
∑ Energy (kJ)	40.3 (29.1/54.4)	40.1 (29.8/45.4)	.741	42.8 (30.1/47.3)	38.3 (26.5/43.9)	.441	.786
Ø Power (W)	33.6 (31.5/35.3)	33.5 (31.6/35.4)	.889	33.6 (29.6/34.1)	36.2 (30.0/45.1)	.072	.081
Ø CF (g)	18.6 (17.0/20.1)	18.4 (17.1/18.9)	.643	17.9 (16.6/18.9)	18.4 (16.4/19.8)	.678	.216
Ø LSI (-)	5.2 (4.9/5.4)	5.2 (4.9/5.4)	.866	5.5 (5.2/5.8)	5.4 (4.9/6.0)	.912	<.001

The term “any Injury” summarizes structural injury (detected by esophagoscopy) and neuropathic alterations (detected by electrogastrography). Ablation parameters are given for the LAPW. Total energy applied (∑ Energy) did not differ. Omission of RF-induced esophageal temperature alarms resulted in a significant reduction in procedural duration. Significant differences are given in bold numbers. Statistical analyses compare (1) patients with/without (w/wo) any esophageal injury, and (2) patients with/without active esophageal cooling (study cohort vs control cohort).

CF = contact force; LA = left atrial; LAPW = left atrial posterior wall; LSI = lesion size index; RF = radiofrequency current.

^∗Procedural duration from vascular access to removal of sheaths, thus including insertion of the cooling or temperature monitoring probe.

Gastrointestinal workup following PVI with active esophageal cooling showed mucosal ELs in 3 of 64 patients (4.7%, all erythema [Kansas City Classification² grade I]), new-onset food retention in 10 of 64 patients (15.6%), and ablation-induced vagal nerve injury in 9 of 64 patients (14.1%). (Figure 2, Supplemental Table 1) Ablation-induced mucosal lesions were not associated with any specific symptoms. There was no association between vagal nerve injury (neuropathic pattern in EGG) and food retention; only 1 patient showed both. A total of 8 of 19 patients with vagal nerve injury reported bloating and abdominal distension which resolved within 4 weeks. Of the LET-monitored control cohort 6 of 52 patients (11.5%) showed mucosal ELs (Kansas City Classification grade I), 14 (26.9%) new-onset food retention, and 15 (28.8%) neuropathic patterns in EGG indicating ablation-induced vagal nerve injury. (Figure 2, Supplemental Table 1) The rate of any esophageal injury per patient was significantly reduced by esophageal cooling which decreased the rate of injury by a factor of 0.51 ([95% confidence interval 0.30, 0.86]; P = .0142).

Figure 2.

Upper gastrointestinal and vagal nerve injury with vs. without esophageal cooling. Percentage of esophageal lesions (ELs), new-onset food retention, and vagal nerve injury (VNI) detected by electrogastrography (EGG) with esophageal cooling (study-group) in green bars compared with the control group (with esophageal temperature monitoring) in grey bars. Each of the different manifestations of esophageal injury (ELs, food retention, and vagal nerve injury) is reduced by esophageal cooling but did not reach statistical significance independently. Using quasi-Poisson-regression, the rate of (any) esophageal injury per patient was statistically significantly reduced (P = .014) with esophageal cooling. Data are given in Supplemental Table 1.

Both with or without esophageal cooling, vagal nerve injury was predominantly detected by gastric arrhythmia rather than a loss in power increase following RF-PVI. (Figure 3, Supplemental Table 2) In the cooling-cohort, pre-existing esophageal vulnerability (eg, esophagitis, erosion, Barrett-mucosa)¹⁶ was seen in 18 patients but this was not associated with an increased likelihood of ablation-induced peri-esophageal injury.

Figure 3.

EGG-parameters before and after ablation (study cohort). The boxplots compare the results of EGG parameters without (green) and with (red) neuropathic patterns) in patients with esophageal cooling, and the differences between post-procedural (dark color) and pre-procedural EGG (bright color). EGG addresses a functional state, thus, each patient has to serve as his or her own control. The ablation-induced effect on these parameters for patients with vs without vagal nerve injury is shown by dark boxes compared with the light ones. Data are given in Supplemental Table 2. The boxes show the interquartile range (with a median line), the arms the maximum values without the outliers (dots). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001 EGG = electrogastrography; PVI = pulmonary vein isolation.

With regard to adverse events, the insertion of the cooling probe was associated twice with short-term malposition in the trachea (repositioned immediately) resulting in coughing. Post-interventional endoscopy showed mechanical abrasion in 5 of 64 patients deemed because of the transesophageal echocardiography probe (in line with prior studies²¹). Cooling did not induce systemic hypothermia. Body temperature declined by 1.3 ± 0.6 K within 73.6 ± 27.6 minutes of active esophageal cooling. No patient reported discomfort associated with esophageal cooling or the probe itself. During follow-up, no patient died or reported signs of delayed esophageal injury.

Scheduled and additional symptom-driven ECG and Holter-ECG recordings (identically in both cohorts), hospital readmissions, and primary care physician records identified arrhythmia relapses (AF, atrial flutter, or atrial tachycardia) in 11 of 64 patients (17%) in a cohort with predominantly persistent AF (mean follow-up period 259 [range 64–435] days), similar to the rate seen in the LET-monitored cohort.

Discussion

PVI by RF ablation is still the most common method in the interventional treatment of AF. Although complication rates overall have declined, the rate of AEF has remained unchanged, and strategies to avoid ablation-induced esophageal damage have failed so far.¹¹

To the best of our knowledge, the reported results are the first data addressing the protective effect of active esophageal cooling on peri-esophageal tissue and the vagal nerve plexus.^17,22,23 Active esophageal cooling resulted in a significant reduction of overall esophageal injury. These subgroups of esophageal injury included a trend for reduced vagal nerve injury (by 51%), as well as a reduction of mucosal lesions (by 59%), and new-onset food retention (by 42%). Food retention (the endoscopic finding associated with gastric hypomotility) poorly overlapped with vagal nerve injury detected by EGG. Thus, vagal nerve injury is likely only part of the pathophysiology of food retention. Total energy applied at the LAPW in the study cohort was not different to the control cohort, thus excluding an impact of ablation settings on the rate of peri-esophageal injury.

Pathophysiology

The pathophysiology of EL progression to fistula is complex and incompletely understood. Vagal nerve injury leads to gastric motility disorders and gives rise to gastro-esophageal reflux, thus perpetuating inflammation and impairing EL healing. Inflammatory cascades activated by tissue-heating are responsible for the progression and extension of heat-induced injury to initially unaffected tissue⁴ and are thought to play a major role in EL progression to fistula.³ In contrast, cooling downregulates mediators of inflammation and has been shown to salvage tissue from burn damage.²⁴

Previous studies and protective efforts

In more than 80,000 RF-PVI procedures worldwide, active esophageal cooling was associated with a reduction of the risk of AEF. In an initial report of 20,000 procedures, there was no AEF.²³ In contrast, esophageal temperature monitoring has failed to reduce ELs.^10,11 However, because of the very rare incidence of AEF (recently estimated as 0.2%),²⁵ studies investigating the effect of esophageal temperature monitoring or active esophageal cooling had to focus on esophageal mucosal lesions (assessed by upper gastrointestinal endoscopy) as a surrogate marker.¹³ Although mucosal lesions are a likely prerequisite of AEF,² fistula is the real target of the preventive efforts, and other potentially relevant contributors of peri-esophageal injury get neglected when the surrogate marker mucosal lesions are used as the end point. So far, esophageal cooling has been studied for its preventive effect regarding mucosal lesions but not for vagal nerve injury.

Thermodynamics

The esophageal mucosa is continuously in close contact with the cooling probe and thus, the direct heat impact of the ablation energy is significantly reduced in the esophageal wall (where it matters).^22,26 In contrast, the thermocouples of the LET probe are positioned at some distance to the heating source and, therefore, may even fail to detect a temperature rise because of the misalignment between the ablation spot and the monitoring probe.²⁷ In contrast, the heating effect of RF ablation declines with the distance from the ablation spot to the esophagus, and theoretically, the effect of active esophageal cooling declines from the mucosa to the peri-esophageal tissue.^3,28 In contrast to our findings in the control cohort with RF-PVI,⁸ there was no association between pre-existing esophageal vulnerability and ablation-induced injury in the presence of active esophageal cooling, possibly because of cooling-induced inhibition of an inflammatory cascade in vulnerable tissue.^3,4 If so, patients with pre-existing inflammation may receive additional benefits from active esophageal cooling.

There have been initial concerns that active esophageal cooling itself might have a harmful effect on the vagal nerve plexus because of its close proximity (and lack of insulation by the fibrous pericardium). However, cooling the esophagus in patients undergoing RF-PVI with a probe at 4 °C resulted in an overall reduction of vagal nerve injury in our study.

Theoretically, cooling in the esophagus close to the ablation site at the LAPW might be associated with lower ablation lesion efficacy in the LA; however, insulating effects from the pericardium and pericardial fat appear to minimize the extension of the effects of esophageal cooling on the atrial myocardium.²² In this study, freedom from arrhythmia recurrence (as the best surrogate for transmurality of ablation lesions) was not different compared with the control cohort with LET monitoring. Other studies have shown improved freedom from atrial arrhythmia during mid-term follow-up after PVI with active esophageal cooling,²⁹ presumed because of improved continuity and transmurality of RF ablation lesions³⁰ in the absence of interruptions of RF-application because of LET-rises (with progressive edema formation at the ablation site).

Previously reported adverse events with active esophageal cooling (esophageal hematoma from introduction of the cooling probe and systemic hypothermia resulting in hypotension)³¹ were not observed in this study. Pre- and post-procedural endoscopy excluded mucosal damage caused by the soft cooling probe. Uninterrupted cooling for up to 2 hours did not result in a reduction of body temperature exceeding 2.5 K.

Clinical considerations and outlook

Recent advances in ablation technology may not share the risk of AEF.³² However, new risk profiles related to pulsed field ablation (eg, hemolysis and acute renal failure),³³ and residual thermal effects of pulsed field ablation^34,35 need to be resolved. The extensive experience with well-established RF ablation remains an asset, and active esophageal cooling may further enhance safety by reducing the likelihood of potentially lethal AEF in thermic RF-PVI. With esophageal cooling (omitting esophageal temperature monitoring), the PVI procedure and the operator are not burdened by numerous alarms resulting from LET-rises, resulting in a streamlined and more efficient procedural workflow. Alternative options for esophageal protection, for example, deviation probes,³⁶ have not been assessed with regard to peri-esophageal vagal nerve injury, and those devices do not necessarily also alter the course of the vagal plexus. To the best of our knowledge, this study is the first to report use of active esophageal cooling in PVI under sedation with unassisted respiration, demonstrating an additional benefit of safety and effectiveness in patients without the need for general anesthesia.

Limitations

This is not a randomized clinical trial; however, patients were treated by the same clinical team, undergoing the same clinical workup and treatment, without any change other than the modality of esophageal protection. As such, a significant systematic bias altering the results is unlikely.

EGG is a technique not typically used in daily clinical practice, but nevertheless is a method that records the functional state of the gastro-intestinal tract with good overall diagnostic accuracy.¹⁹ Although the use of EGG is labor-intensive, it is not unlike other procedures such as endoscopy that provide valuable data when warranted, and for this study, its use provides useful insight. Brief asymptomatic AF recurrences were beyond the scope of this specific study.

Conclusion

In RF-PVI, active esophageal cooling reduces ablation-induced vagal nerve injury and overall peri-esophageal injury.