Abstract
Background Esophageal thermal injury is a risk of ablation of the posterior left atrium despite various devices utilized to date.
Objective Evaluate the potential of a commercially-available esophageal cooling device to provide esophageal protection during left atrial catheter ablation.
Methods In this pilot study, we randomized 6 patients undergoing catheter ablation for atrial fibrillation. Three patients received standard of care for our site (use of a single-sensor temperature probe, with adjunct iced-water instillation for any temperature increases >1°C). Three patients received standard ablation after placement of the esophageal cooling device using a circulating water temperature of 4°C. All patients underwent transesophageal echocardiogram (TEE) and esophagogastroduodenoscopy (EGD) on the day prior to ablation followed by EGD on the day after.
Results In the 3 control patients, one had no evidence of esophageal mucosal damage, one had diffuse sloughing of the esophageal mucosa and multiple ulcerations, and one had a superficial ulcer with large clot. Both patients with lesions were classified as Zargar 2a. In the 3 patients treated with the cooling device, one had no evidence of esophageal mucosal damage, one had esophageal erythema (Zargar 1), and one had a solitary Zargar 2a lesion. At 3-month follow-up, 1 patient in each group had recurrence of atrial fibrillation.
Conclusions The extent of esophageal injury was less severe with a commercially available esophageal cooling device than with reactive instillation of ice-cold water. This pilot study supports further evaluation with a larger clinical trial.
Introduction
The treatment of atrial fibrillation via pulmonary vein isolation using radiofrequency (RF) ablation has well documented success. A rare complication of this procedure is atrial-esophageal fistula (AEF) formation.[1,2] A precursor to AEF involves esophageal submucosal and possibly mucosal injury.[3] This damage is delivered by heating of the posterior aspect of the left atrium to create a transmural infarction that may inadvertently cause further thermal injury to esophageal tissue. Esophageal mucosal injury occurs in up to 30-50% of patients undergoing ablation.[4]
Attempts to protect the esophagus have been evaluated including mechanical displacement of the esophagus as well as multiple cooling modalities. Cooling approaches that have been evaluated include ice lavage, a cooling sac (EPSac, RossHart Technologies Inc.), and multiple types of cooling balloons. Many of these devices have been tested in bench models and animals with limited human use.[5-12] However, these earlier attempts at esophageal protection have not completely eliminated the risk of esophageal damage.
One commonly used approach to esophageal protection involves luminal esophageal temperature (LET) monitoring with adjustment in ablation location and timing based on temperature change. However, the efficacy of this approach remains uncertain.[13] Moreover, it has been suggested that the LET probe also contributes to thermal injury.[14] For this reason, we performed a pilot study of a commercially available esophageal cooling device currently utilized for whole-body temperature management, and shown to have protective effects in animal and mathematical models, as well as in recently presented clinical data.[15,16] The primary objective was to evaluate the potential to utilize this device for the reduction in the occurrence and severity of esophageal injury in patients undergoing RF ablation.
Methods
Study population
Six adults diagnosed with an atrial arrhythmia were included in this blinded (patient and outcome assessor), randomized controlled pilot study. Patients considered for ablation of the posterior wall (non-pregnant, between age 18-90 years, and without prior esophageal damage or disease) were asked to participate in the study after written informed consent. The study was approved by the hospital’s institutional review board, and the study was registered on Clinicaltrials.gov with identifier #NCT03481023.
Permutated block randomization was used for group designation. The study included 5 male and 1 female subjects aged 55 to 71 years. Two of the patients had mild reduction in left ventricular ejection fraction (LVEF) with the remainder of the patients having normal LVEF. Pre-procedural diagnoses included paroxysmal atrial fibrillation (5 of 6), typical atrial flutter (1 of 6), and atypical atrial flutter (2 of 6). All patients had failed at least 1 antiarrhythmic agent prior to the procedure. An outpatient visit was arranged at 3 months following ablation to assess for adverse events and to arrange further evaluation of atrial arrhythmia recurrence. (Table 1)
Intervention
Patients received a transesophageal echocardiogram (TEE) the day prior to ablation, which was coupled with endoscopic evaluation including esophagogastroduodenoscopy (EGD). An independent gastroenterologist blinded to the patient’s assignment graded esophageal mucosal damage. We used Zargar’s Modified Endoscopic Classification Scheme to provide objective quantification of mucosal changes (Table 2). This classification scheme expands on the customary endoscopic classification of burns (grades 0 to 3) by subdividing grade 2 burns into 2a and 2b, and grade 3 burns into 3a and 3b for prognostic and therapeutic implications.
All patients had general anesthesia as is customary for this institution. All ablations were performed by the same electrophysiologist with the choice of ablation including any combination of pulmonary vein isolation, true left atrial roof line, mitral line, cavo-tricuspid isthmus, and complex fractionated atrial electrogram ablation as was clinically indicated. Patients received a single-sensor temperature probe placed in the esophagus, as well as a standard nasogastric tube placed to the depth of the left atrium. The temperature readings were monitored by lab staff during ablations on the posterior wall. In the event that temperature rose by more than 1°C from baseline, ablation would cease, and the anesthesiologist was instructed to instill ice-cold water into the nasogastric tube in 10-20 mL aliquots. Ablation would then resume after equilibration of temperature readings to patient baseline. Ablation was carried out with 35 watts of energy at 55 degree temperature on the posterior wall and 50 watts of energy and 55 degree temperature on the anterior wall using a 4mm tip RF catheter. The RF catheters were non-contact-force sensing and non-irrigated.
The treatment group included the same pre-operative assessment with TEE followed by EGD on the day prior to ablation, and post-operative care and assessment was also identical to the control group. The treatment group patients received placement of a commercially available esophageal cooling device (figure 1) into the esophagus. This device provides a closed-circuit flow of water through a multi-channel cylindrical silicone tube placed in the esophagus analogously to a standard orogastric tube, heating or cooling a patient through conductive heat transfer across the esophagus, and convective heat transfer through the device.[17] The device was connected to a heat exchange unit (Blanketrol III Hyper-Hypothermia System, Gentherm Medical, Cincinnati, OH) which allows for large volume flow of temperature regulated distilled water at a rate of 136 L/hour. Just prior to ablation the coolant was set to 4°C, and this temperature was maintained during ablation on the posterior wall. The patient’s core temperature was monitored via a rectal temperature probe.
All patients were observed in the hospital overnight with repeat EGD performed on the morning following the ablation procedure. Follow-up care included a 1-2 week assessment as well as a 3 month visit. All patients had an event monitor or loop recorder between 3-6 months following ablation.
Results
All of the patients had normal esophageal mucosa immediately following TEE on the day prior to ablation. One patient was incidentally noted to have Barret’s esophagus which required long-term monitoring. In the 3 control patients, all had multiple instillations of ice-cold water for temperature excursions above 1 C during posterior wall ablation. Of these 3 patients, one had no evidence of esophageal mucosal damage, one had diffuse sloughing of the esophageal mucosa and multiple ulcerations, and one had a superficial ulcer with large clot. Example EGD images of control patient lesions are shown in Figure 2. Both patients with lesions were classified as Zargar 2a (Table 3).
In the 3 patients treated with the esophageal cooling device, one had no evidence of esophageal mucosal damage, one had esophageal erythema (Zargar 1), and one had a solitary Zargar 2a lesion. There was no significant difference in the total ablation time or posterior wall ablation between the groups, as seen in Table 4. It is notable that the patient with the most extensive posterior wall ablation in the treatment group did not have any esophageal injury (Table 3). Example EGD images of the treatment arm patients are shown in Figure 3. Placement of the esophageal cooling device was straightforward and did not interfere with ablation, allowing the procedure to continue without interruption for temperature overshoot. Moreover, there was no need for fluoroscopy use since the cooling device stays placed to the depth of the stomach without need for repositioning during the procedure.
The core temperature of all subjects was monitored via rectal temperature sensor. There was minimal systemic cooling effect noted. The device was additionally used as an aid in rewarming the patient at the conclusion of the procedure by setting the system to a warming mode. As commonly described, the systemic temperature of the control patients was also noted to reduce with general anesthesia despite the use of forced-air blanket warming.
Patients were evaluated via event monitor or loop recorder. Evaluation at between 3-6 months after the ablation procedure showed recurrence of atrial fibrillation in one control patient and one treatment arm patient.
Discussion
In this first randomized, controlled pilot study of a commercially available esophageal cooling device utilized for esophageal protection during left atrial ablation for the treatment of atrial fibrillation, we found fewer high-grade lesions compared to control patients receiving intermittent instillation of ice-cold water. Specifically, the treatment arm cohort included a single ulcer in one patient and an erythematous patch in another patient. The control group included two patients with esophageal lesions with a higher degree of injury, diffuse sloughing, and larger ulcerations.
Although not powered for statistical significance, the potential reduction of lesion incidence and severity may offer a new option for esophageal protection and the prevention of AEF. In addition, because stopping ablation for temperature rises in the esophagus did not occur, and because repositioning of the traditionally-used temperature probe was not required, an improvement in workflow, with a corresponding reduction in fluoroscopy usage, is suggested.
Various approaches have been developed to attempt esophageal protection during RF ablation, including LET monitoring, cooling, and deviating the esophagus. Nevertheless, currently available discrete sensor probes, whether single or multiple, do not appear to significantly reduce injury rates, and there is a potential for esophageal harm with esophageal deviation [18-20] In contrast, multiple earlier studies have evaluated esophageal cooling using a variety of techniques, with all but one suggesting potential benefit.[5-12] The availability now of a commercial device currently on the market for whole-body temperature modulation offers the chance to further study this modality, which in our pilot data, appears to offer potential. A number of larger clinical studies of this approach are ongoing: Improving Oesophageal Protection During AF Ablation (IMPACT) - NCT03819946, Esophageal Cooling for AF Ablation (eCoolAF) - NCT03691571, and Esophageal Damage Protection During Pulmonary Vein Ablation. Pilot Study. - NCT03832959.
Limitations
We were not able to blind the physician performing the ablation, and as such, differences in ablation technique may have occurred. Nevertheless, because cessation of ablation occurred regularly in the control arm after elevated temperature probe readings, but did not occur in the treatment arm, it is likely that the actual regional energy deposition over time was greater in the treatment arm. Our standard practice also utilizes non-contact-force sensing, non-irrigated catheters. However, non-irrigated and non-contact-force sensing catheters are used by 30% of the writing group of the 2017 HRS-EHRA-ECAS-APHRS-SOLAECE Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation, so generalizability is not precluded.[21] We attempted to include assessment of the submucosal tissue architecture using endoscopic ultrasound (EUS) to identify tissue abnormalities which may be missed on EGD; however, technical challenges rendered this uninformative. These technical challenges included difficulties in determining exact longitudinal distance of the ultrasound probe relative to EGD scope, and uncertainties in determining abnormal findings, since EUS has not commonly been used for this particular purpose.
Conclusions
The extent of esophageal injury was less severe with a commercially-available esophageal cooling device than with periodic manual instillation of ice-cold water, and there was no reduction in ablation efficacy. Moreover, the efficiency of ablation procedures was improved by eliminating stoppage for temperature overshoot or temperature probe repositioning. This pilot study supports further evaluation with a larger clinical trial.
Data Availability
Available data will be provided in the manuscript without external links.
Footnotes
Funding: Riverside Medical Center Research Department. The cooling catheters were provided by Attune Medical but no funding provided.
Conflict of interest: None.