958

Multi-Sensor Esophageal Temperature Probe Used DuringRadiofrequency Ablation for Atrial Fibrillation is Associated

with Increased Intraluminal Temperature Detection andIncreased Risk of Esophageal Injury Compared to

Single-Sensor ProbeBRETT J. CARROLL, M.D.,∗ FERNANDO M. CONTRERAS-VALDES, M.D.,∗

E. KEVIN HEIST, M.D., Ph.D.,† CONOR D. BARRETT, M.D.,† STEPHAN B. DANIK, M.D.,†JEREMY N. RUSKIN, M.D.,† and MOUSSA MANSOUR, M.D.†

From the ∗Department of Medicine; and †MGH Heart Center, Cardiac Arrhythmia Service, Massachusetts General Hospital, HarvardMedical School, Boston, Massachusetts, USA

Single vs Multi-Sensor ETM in AF Ablation. Background: Radiofrequency (RF) ablation inthe posterior left atrium has risk of thermal injury to the adjacent esophagus. Increased intraluminalesophageal temperature has been correlated with risk of esophageal injury. The objective of this study wasto compare esophageal temperature monitoring (ETM) using a multi-sensor temperature probe with 12sensors to a single-sensor probe during catheter ablation for atrial fibrillation (AF).

Methods and Results: We compared the detection of intraluminal esophageal temperature rises in 543patients undergoing RF ablation for AF with ETM. Esophageal endoscopy (EGD) was performed on allpatients with maximum esophageal temperature ≥39◦C. Esophageal lesions were classified by severity asmild or severe ulcerations. Four hundred fifty-five patients underwent RF ablation with single-sensor ETMand 88 patients with multi-sensor ETM. Thirty-nine percent of patients with single-sensor versus 75% withmulti-sensor ETM reached a maximum detected esophageal temperature ≥39◦C (P < 0.0001). Esophagealinjury was detected by EGD in 29% of patients with maximum temperature ≥39◦C by single-sensor versus46% of patients with multi-sensor ETM (P = 0.021). Thirty-nine percent of patients with lesions in thesingle-sensor probe group had severe ulcerations compared to 33% of patients in the multi-sensor probegroup (P = 0.641).

Conclusions: Intraluminal esophageal temperature ≥39◦C is detected more frequently by the multi-sensor temperature probe versus the single-sensor probe, with more frequent esophageal injury and withcomparable severity of injury. Despite detecting esophageal temperature rises in more patients, the multi-sensor probe may not have any measurable benefit compared to a single-sensor probe. (J CardiovascElectrophysiol, Vol. 24, pp. 958-964, September 2013)

atrial fibrillation, catheter ablation, complications, electrophysiology, esophageal injury, esophageal temperaturemonitoring, pulmonary vein isolation

Dr. Heist has served as a consultant for Biosense Webster, Boston Scientific,Sanofi-Aventis, Sorin, and St. Jude Medical, and has received research grantsfrom Boston Scientific and Sorin. Dr. Ruskin has served as a consultantfor Advanced Medical Education, Astellas/Cardiome, Atricure, BiosenseWebster, Bristol-Myers Squibb, Medtronic, Pfizer, Portola, Sanofi-Aventis,and Third Rock Ventures; has received fellowship support from BiosenseWebster, Boston Scientific, Medtronic, and St. Jude Medical; has served onthe clinical oversight committee for CardioFocus; has served on the scientificadvisory board for CardioInsight and InfoBionic; has equity with Portola;and has served on the scientific steering committee for Pfizer. Dr. Mansourhas received research grants from Biosense Webster, Boston Scientific, St.Jude Medical, MC 10, Voyage Medical, and Rhythmia Medical, and hasserved as a consultant for Biosense Webster and St. Jude Medical. Otherauthors: No disclosures.

Address for correspondence: Moussa Mansour, M.D., 55 Fruit StreetGRB-109, Boston, MA 02114, USA. Fax: 617-724-1241; E-mail:mmansour@partners.org

Manuscript received 22 December 2012; Revised manuscript received 13March 2013; Accepted for publication 16 April 2013.

doi: 10.1111/jce.12180

Introduction

Radiofrequency (RF) ablation is one of the primary treat-ments for symptomatic, drug-refractory atrial fibrillation(AF). The procedure attempts to isolate the pulmonary veins(PV) and often the posterior wall of the left atrium (LA).1

Although the goal is ablation of only the atrial myocardium,damage to adjacent structures is a well-recognized adverseconsequence of ablation. Such damage can occur to the ad-jacent esophagus, which often is in direct contact with theposterior LA.2,3 Studies evaluating the esophagus with post-procedure esophageal endoscopy (EGD) have shown a widerange of incidence of esophageal thermal injury from 0 to47%.4-10 It is believed that these thermal injuries have po-tential to progress to formation of a fistula between the LAand esophagus.11 Atrioesophageal (AE) fistula is rare withreported incidence of 0.03–0.2%; however, the mortality canbe upwards of 75%.12,13

Several strategies have been deployed to decrease therisk of esophageal injury, including limitation of powerused along the areas closest to the esophagus,9 evaluation

Carroll et al. Single vs Multi-Sensor ETM in AF Ablation 959

Figure 1. The single-sensor probe (A andC) has a lone temperature sensor at thetip of the probe, approximately 3 mm inwidth. The multi-sensor probe (B and D)has a sinusoidal shape, approximately 18mm in width, with 12 different sensors.

of the anatomic relationship of esophagus to LA withpreprocedural or real-time imaging,14 displacement of theesophagus,15 and cooling of the esophagus.16 The effective-ness and utilization of such techniques is variable.1 Intra-luminal esophageal temperature monitoring (ETM) is an-other measure frequently used during AF ablation proce-dures. Temperatures detected can be used to help guidethe ablation in areas closest to the esophagus in attempt tolimit heat exposure to the esophagus. Despite this approach,esophageal ulceration continues to occur.6,7,10 ETM is likelylimited by 3 main factors, specifically, difficulty in maintain-ing close proximity of the probe to the catheter given thedynamic nature of the esophagus and constant manipulationof the catheter,17 unpredictable distance of the temperatureprobe from the site of greatest esophageal heating, and theproposed mechanism of esophageal injury from adventitiato lumen. The majority of ETM probes consist of a sin-gle temperature sensor, with some using up to 3 sensors.8,10

A particularly limiting factor with a single-sensor probe isproximity of the sensor to the area of ablation with needfor constant craniocaudal manipulation, but with minimalability to manipulate medially or laterally. A new temper-ature probe has been developed with 12 temperature sen-sors with a 2-dimensional, sinusoidal design to allow fora greater area of temperature monitoring (see Fig. 1). Thisstudy aimed to evaluate the ability of this multi-sensor ETMprobe to detect intraluminal esophageal temperature rise andthe resultant incidence of thermal injury and severity ofesophageal damage in comparison to a traditional single-sensor probe.

Methods

Patient Population

Patients undergoing RF catheter ablation for symptomatic,drug-refractory AF from January 2009 to June 2012 with useof ETM were enrolled in the study. Patients who underwentablation from January 2009 to February 2012 had ETM witha single-sensor probe. Those that underwent ablation fromFebruary 2012 to June 2012 had ETM with a multi-sensorprobe. Patients were compared based on type of ETM used.

RF Catheter Ablation Procedure

Each patient in this study cohort underwent magnetic res-onance or computed tomography for evaluation of LA andPV anatomy. General endotracheal anesthesia was used forall procedures. The ablation procedure was performed asfollows: intravascular introducer sheaths were placed in thefemoral veins. A standard decapolar catheter (Webster CSCatheter, Biosense Webster, Inc, Diamond Bar, CA, USA)was placed in the coronary sinus and used for pacing andrecording. Two transseptal punctures were performed underfluoroscopy and intracardiac echo (AcuNav, Siemens, Moun-tain View, CA, USA) guidance. A long 8-Fr (SL-1, St. JudeMedical, St. Paul, MN, USA) and a deflectable sheath (Ag-ilis, St. Jude Medical, St. Paul, MN, USA) were advancedto the LA. All procedures were performed on uninterruptedwarfarin. If the INR was found to be subtherapeutic, trans-esophageal echocardiogram was performed to rule out intra-cardiac thrombus. Intravenous heparin was administered in

960 Journal of Cardiovascular Electrophysiology Vol. 24, No. 9, September 2013

bolus prior to transseptal punctures, followed by an infusionto maintain an activated clotting time of 300–350 seconds.Bipolar electrograms were recorded at a bandpass of 30–500Hz (PruckaCardioLab IT System, GE Healthcare, Milwau-kee, WI, USA). A circular mapping catheter (Lasso, BiosenseWebster, Inc) was used to create the electroanatomical mapsand confirm PV isolation. The electroanatomical maps of theLA and the PVs were created and reconstructed by usingthe Carto system (Biosense Webster, Inc) or the EnSiteNavXsystem (St. Jude Medical). An externally irrigated 3.5-mm-tip quadripolar catheter (Thermocool, Biosense Webster, Inc)was used for mapping and ablation. RF applications were de-livered circumferentially to the antral regions of all 4 PVs inthe majority of patients. Ipsilateral PVs were isolated witha single ablation ring when possible. In some cases, addi-tional lesions were performed in the posterior LA as well asother sites in the LA and RA as indicated by each clinicalscenario, including box lesions in some. A Stockert gener-ator (Biosense Webster, Inc) was set to deliver RF lesions.Catheter temperature, power, and impedance were recordedwith each RF application. When creating lesions in the pos-terior LA, power was limited to 25 W and to 30 seconds.

Temperature Monitoring

In the control group, temperature was monitored witha single-sensor, 9 French esophageal temperature probe(Acoustascope, Smiths Medical ASD, Inc, Keene, NH, USA)(Fig. 1A and C). The probe was inserted orally and ad-vanced posterior to the LA, under fluoroscopic guidance.The anatomic course of the probe was constantly visualizedby fluoroscopy, and its craniocaudal position was continu-ously adjusted to the level of the ablation catheter tip insidethe LA during posterior LA and PV ablation. The accuracyof the thermistor is reported to be plus or minus 0.3◦C bythe manufacturer. The temperature detecting portion of theprobe is approximately 3 mm in width. The single sensorprobe continuously measured temperature and was linked tothe PruckaCardioLab IT system by using a Tram module(Tram-rac 4A module, GE Medical Systems, Waukesha, WI,USA), which displayed real-time temperature. A 10 Frenchesophageal temperature probe (Circa S-Cath, Circa Scien-tific, Park City, UT, USA) was used to continuously recordthe intraluminal esophageal temperature in the multi-sensorprobe group. The probe was inserted orally under fluoro-scopic guidance with use of a straightening stylet and ad-vanced into the esophagus, posterior to the LA. The probehas a 2-dimensional sinusoidal shape, with 12 sensors thatreport temperature readings 20 times per second (Fig. 1B andD). The accuracy of the thermistor is reported to be plus orminus 0.3◦C by the manufacturer. The temperature detectingportion of the probe is approximately 18 mm in width. Themulti-sensor probe was connected to the Circa TemperatureMonitoring System (CS-1000 Circa Temperature Monitor-ing System, Circa Scientific, Park City, UT, USA) with 50millisecond refresh rate and continuous maximum tempera-ture display. Energy delivery was terminated if at any timethe esophageal temperature increased to 38◦C and resumedonly after the temperature returned to that patient’s baseline.Ablation would frequently be moved closer to or further fromthe PV ostia if heating was observed (using fluoroscopy tomove the ablation catheter further from the esophagus), alongwith decrease in power and/or duration of ablation lesions. It

was not infrequent that intraluminal esophageal temperaturewould continue to rise above threshold and reach 39◦C orhigher despite termination of energy delivery at 38◦C.

Endoscopic Evaluation

EGD was performed the day after the procedure in pa-tients with esophageal heating to ≥39◦C.18,19 Patients withesophageal injury were treated with omeprazole 40 mg BID,and their diet changed to pureed consistency until the ulcerwas healed. For patients without esophageal heating aboveour threshold, no EGD was performed and omeprazole wasempirically given at 40 mg BID for 7 days, followed by 20mg daily for 3 additional weeks. The same protocol was alsofollowed for patients without esophageal injury evident onEGD. For patients with any evidence of esophageal injury,repeat EGD was obtained approximately 10–14 days after theinitial endoscopy. If damage was still present, further EGDswere performed at approximately 2-week intervals until heal-ing was documented. Lesions were independently classifiedby gastroenterology and electrophysiology physicians whowere blinded to the type of ETM probe used. Esophageallesions were classified as mild or severe lesions.18 Lesionswere classified as severe if they exhibited evidence of mu-cosal ulceration or submucosal hemorrhage.

Follow-Up

Patients were typically discharged from the hospital on theday following the procedure. Anticoagulation was continuedfor at least 3 months after the ablation. Antiarrhythmics werestopped after ablation in patients with paroxysmal AF, andmaintained at preablation levels during the first 2 monthsfollowing ablation in patients with persistent AF. Patientswere seen in the outpatient clinic for follow up at 6 weeksand 3 months after the procedure.

Statistical Analysis

Statistical analyses were performed with JMP Pro 10.0.0(SAS Institute Inc., Cary, NC, USA). The association be-tween type of ETM probe used and incidence of temperaturedetection ≥39◦C, thermal injury seen on EGD, and severityof injury were analyzed with the use of a two-by-two contin-gency table using Fisher’s exact test. Patient characteristicsof continuous variables were compared by using Student’st-test, while categorical variables were analyzed with theFisher’s exact test.

Results

Patient Characteristics

A total of 543 AF ablation procedures with ETM wereperformed between January 2009 and June 2012. The studywas nonrandomized, and the multi-sensor probe was usedprimarily in the later portion of this timeframe when it be-came commercially available. Of these patients, 455 patientshad ETM with a single-sensor and 88 with a multi-sensorprobe. The demographic and clinical characteristics of thepatients are summarized in Table 1. The only statisticallysignificant differences between the two groups were greatertotal RF ablation time in the multi-sensor probe group (86.3minutes vs 63.7 minutes, P < 0.0001) and total energy ap-plication in the multi-sensor group (134,765 Watt seconds

Carroll et al. Single vs Multi-Sensor ETM in AF Ablation 961

TABLE 1

Demographic and Clinical Characteristics of Patients at Baseline

Single-Sensor Multi-SensorProbe Probe

(n = 455) (n = 88) P Value

Age 61.3 ± 8.8 61.7 ± 10.1 0.665Gender

Male (%) 72.7 70.4 0.696BMI (kg/m2) 29.5 ± 5.1 28.4 ± 4.7 0.070LV EF

<50% (%) 13.3 10.0 0.281LA A-P diameter (mm) 42.4 ± 6.9 43.8 ± 7.3 0.077Type of AF

Paroxysmal (%) 48 53Persistent (%) 52 47 0.354

Type of anesthesiaGeneral (%) 100 100 1.0

Total RF ablation 63.7 ± 25.7 86.3 ± 31.4 <0.0001time (min)

Total RF lesions 133 ± 53 128 ± 50 0.349Total energy 9.87 ± 3.9 13.48 ± 6.2 <0.0001

application × 104 (Ws)Average power (Watts) 25.6 ± 4.4 25.6 ± 3.6 1.0Box lesion (%) 70 66 0.452

Values are reported as the mean ± standard deviation or number of patients(%). AF = atrial fibrillation; A-P = anteroposterior; BMI = body massindex; EF = ejection fraction; LA = left atrium; LV = left ventricle; RF =radiofrequency; Ws = Watt seconds (Joules).

vs 98,651 Watt seconds, P < 0.0001). There were no sta-tistically significant differences in average power, total RFlesions, or percent of box lesions performed between the twogroups. There was no AE fistula identified in this cohort. AllPVs were determined to be isolated at the end of the ablation.Results of esophageal temperature detection and injury aresummarized in Figure 2.

Esophageal Temperature Monitoring

A maximum intraluminal esophageal temperature ≥39◦Cwas recorded in 75% (66/88) of patients with ETM usingthe multi-sensor probe. Comparatively, a temperature rise≥39◦C occurred in 39% (177/455) of patients with ETM us-ing the single-sensor probe (75% vs 38.9%, P < 0.0001).Within each sensor cohort, there was a significantly greateramount of total RF ablation time, total number of RF lesions,and total energy applied in patients with an esophageal tem-

perature rise ≥39◦ compared to those that did not reach thatthreshold (see Table 2).

Esophageal Evaluation

Among those patients who had an intraluminal esophagealtemperature rise ≥39◦C, an EGD was performed for evalu-ation of possible injury. Evidence of thermal injury to theesophagus was seen in 30 of the 66 patients (45.5%) inmulti-sensor probe group compared to 51 of the 177 patients(28.8%) who had ETM with the single-sensor probe (P =0.021). There were no statistically significant differences inpatient characteristics within each sensor cohort between pa-tients with injury and those without injury (see Table 3).

There was also no statistically significant difference inthe severity of lesions detected between the 2 groups. Thirty-three percent of lesions in the multi-sensor group were cate-gorized as severe lesions compared to 39% of those lesionsin the single-sensor group (P = 0.64) (see Fig. 3)

Discussion

This study shows a significantly increased rate of intra-luminal esophageal temperature rise to ≥39◦C detected us-ing a multi-sensor probe compared to a single-sensor probe.Interestingly, despite an increased rate of temperature risedetection, there was a significantly increased incidence ofesophageal injury with the multi-sensor probe.

Intraluminal Esophageal Temperature Monitoring

Intraluminal ETM may be used as a possible surrogate forrisk of esophageal injury, though varying data exists regard-ing correlation between maximal temperature detected anddevelopment of injury.5,8,10,20 ETM may reduce the risk ofesophageal ulceration,5,8 with Singh et al. showing an 83%relative risk reduction.10 However, esophageal thermal in-jury can occur at rates up to 40% despite its use.7 Luminalesophageal temperature likely underestimates the risk of ther-mal injury, with significantly increased temperatures foundat the adventitial surface of the esophagus that do not neces-sarily correlate with intraluminal esophageal temperatures.21

Conductive heating is the most commonly accepted theoryfor the mechanism of injury.8,19,21-24 Damage likely occursfrom the adventitia of the esophagus inwards to the lumen,with muscle absorbing much of the heat.22,23 Furthermore,esophageal injury may not be related to the direct effect of

Figure 2. The study included a total of543 AF ablations. The frequency of tem-perature rise ≥39◦C and esophageal in-jury, along with comparison of severity ofinjury, is summarized in this flow chart.AF = atrial fibrillation; EGD =esophageal endoscopy; T = maximumintraluminal temperature detected.

962 Journal of Cardiovascular Electrophysiology Vol. 24, No. 9, September 2013

TABLE 2

Demographic and Clinical Characteristics of Patients with Temperature Detection ≥39◦C

Single-Sensor Probe Multi-Sensor Probe

Temp ≥39◦C Temp <39◦C Temp ≥39◦C Temp <39◦C(n = 177) (n = 278) P Value (n = 66) (n = 22) P Value

Age 60.6 ± 8.7 61.7 ± 8.8 0.202 61.9 ± 10.1 61.1 ± 10.5 0.745Gender

Male (%) 76.8 70.1 0.131 68.1 77.2 0.591BMI (kg/m2) 29.4 ± 5.0 29.5 ± 5.2 0.745 28.6 ± 4.8 27.9 ± 4.5 0.523LV EF

<50% (%) 12.9 11.8 0.77 9.1 4.5 0.675LA A-P diameter (mm) 41.7 ± 6.4 42.7 ± 7.1 0.102 43.7 ± 7.4 44.3 ± 7.4 0.730Type of AF

Paroxysmal (%) 51.4 45.7 53.0 54.5Persistent (%) 48.6 54.3 0.249 47.0 45.5 1.0

Total RF ablation time (min) 68.1 ± 23.4 61.8 ± 25.9 0.010 95.5 ± 26.4 58.8 ± 29.2 <0.0001Total RF lesions 139 ± 50 129 ± 54 0.051 139 ± 47 92 ± 42 <0.0001Total energy application × 104 (Ws) 10.41 ± 3.6 9.51 ± 4.1 0.017 14.82 ± 5.6 8.63 ± 5.7 0.0001Average power (Watts) 25.6 ± 3.6 25.7 ± 4.9 0.815 25.6 ± 3.1 25.6 ± 4.9 1.0Box lesion (%) 73 68 0.345 70 55 0.206

Values are reported as the mean ± standard deviation or number of patients (%). AF = atrial fibrillation; A-P = anteroposterior; BMI = body mass index;EF = ejection fraction; LA = left atrium; LV = left ventricle; RF = radiofrequency; Ws = Watt seconds (Joules).

TABLE 3

Demographic and Clinical Characteristics of Patients with Injury

Single-Sensor Probe Multi-Sensor Probe

Injury No Injury Injury No Injury(n = 51) (n = 126) P Value (n = 30) (n = 36) P Value

Age 61.7 ± 8.2 60.2 ± 8.8 0.277 59.9 ± 10 63.6 ± 10 0.144Gender

Male (%) 77.0 76.5 1 73.3 63.9 0.44BMI (kg/m2) 28.5 ± 3.9 29.7 ± 5.4 0.137 29.8 ± 5.1 27.6 ± 4.4 0.076LV EF

<50% (%) 13.7 12.7 0.810 13.3 5.6 0.399LA A-P diameter (mm) 41.3 ± 6.3 41.9 ± 6.5 0.543 45.3 ± 8.7 42.3 ± 5.9 0.113Type of AF

Paroxysmal (%) 56.9 49.2 46.7 58.3Persistent (%) 43.1 50.8 0.408 53.3 41.7 0.458

Total RF ablation time (min) 69.6 ± 19.4 67.5 ± 24.8 0.606 98.7 ± 24.5 92.8 ± 28.0 0.371Total RF lesions 141 ± 48 139 ± 51 0.763 148 ± 50 133 ± 43 0.203Total energy application X 104 (Ws) 10.51 ± 3.1 10.37 ± 3.8 0.815 15.62 ± 7.0 14.2 ± 4.3 0.301Average power (Watts) 25.8 ± 1.9 25.5 ± 4.0 0.610 25.9 ± 4.5 25.3 ± 1.2 0.445Box lesion (%) 73 73 1.0 77 64 0.294

Values are reported as the mean ± standard deviation or number of patients (%). AF = atrial fibrillation; A-P = anteroposterior; BMI = body mass index;EF = ejection fraction; LA = left atrium; LV = left ventricle; RF = radiofrequency; Ws = Watt seconds (Joules).

ablation on the tissue alone, but compounded by effects onthe periesophageal vagal nerve and damage to the anterioresophageal arteries.2,19,25 An additional limitation lies in thevariable proximity of the temperature probe to the site of ab-lation. The esophagus is broad and the tip of a single-sensorETM probe may be displaced laterally and/or posteriorly incomparison to the site of ablation, even with craniocaudalmanipulation of the probe.5,10 Animal studies have shownthat a steep gradient exists between areas of highest andlowest luminal temperature,26 with Leite et al. showing ma-nipulation of a deflectable temperature probe was frequentlyrequired to increase sensing ability.4

The use of multi-sensor probes has been employed in anattempt to decrease the variable distance between catheterand probe.8,27 Previous multi-sensor probes studied had 3sensors, were linear in design, and showed high temperaturedetection rates from 56% to 84%.7,19,28 In our study, there

was also a high rate of temperature detection with use ofthe multi-sensor probe, with a rate of 75%, almost doublethat seen with the single-sensor probe (38.9%). This may besecondary to the greater area of temperature detection withmore sensors and a sinusoidal design.

Esophageal Injury

An unexpected finding in our study was a significantly in-creased incidence of esophageal injury with the multi-sensorprobe compared to the single-sensor probe (45.5% vs 28.8%,P = 0.021) among the patients with a temperature rise ≥39◦C.The reason for this increased incidence of injury despite in-creased temperature detection rates is not clear. The onlysignificant difference in patient characteristics was a greateramount of total RF ablation time and total energy applied inthe multi-sensor group, with no difference in average power

Carroll et al. Single vs Multi-Sensor ETM in AF Ablation 963

Figure 3. Esophageal lesions were categorized into 2 groups, mild or se-vere. Mild lesions are characterized by shallow involvement of the mucosaassociated with erythema. Severe lesions generally involve multiple layers ofmucosa with a hemorrhagic appearance and often with fibrinoid material.

used or number of RF lesions. A possible explanation for thisdifference could be that the increased temperature detectionwith the multi-sensor probe altered ablation away from thePV, thus creating a need for increased overall ablation. In-terestingly, there was no significant identifiable difference intotal RF ablation time or total energy applied between thosepatients who had injury and those who did not within eachcohort group (see Table 3). Similar to our findings, 2 priorstudies also showed no association between ablation time ortotal energy applied and esophageal injury. Martinek et al.did not show a significant difference in Watt seconds of totalablation or ablation applied to the posterior wall betweenpatients with and without esophageal injury.9 Yamasakiet al. also did not show a significant difference in total amountor duration of RF energy delivery at the LA posterior wallbetween those with injury and those without.5 Data are notavailable to specify the amount of posterior LA ablation inour study, though there was no significant difference in boxlesions.

One possibility for the increased incidence of injury withthe multi-sensor probe is a change in configuration of theesophagus, thus increasing the width of the esophagus incontact with the LA and/or tenting the esophagus againstthe LA. Another possibility is an interaction between theablation catheter and the multi-sensor probe itself. Denekeet al. found increased esophageal injury among those usingETM compared to no ETM.19 The authors suggested thehypothesis that there was greater conductivity between themetal temperature sensors of the probe and RF application,leading to direct esophageal heating and injury from the tem-perature probe itself. They site unpublished ex vivo data thatshowed an additional thermal effect on large metal alloyswhen placed near an ablation catheter. They also hypothe-size the metal electrodes could have functioned as antennas,altering the electrical field distribution.

Study Limitations

Our study was not randomized and was retrospective. Al-though patient characteristics were not significantly differentbetween groups other than RF ablation time, confoundingvariables may exist that would interfere with our analysis.Patients without esophageal temperature rise ≥39◦C did notundergo endoscopy, and some of these patients may have

therefore had esophageal lesions that were not detected.There is no defined intraluminal esophageal temperature risethreshold below which no esophageal thermal injury mayoccur. Thus, there is the potential to miss possible injury inpatients that did not undergo EGD; however, it should benoted that the same temperature threshold was used in eachgroup. Also, there is a limitation in the categorization of theseverity of lesions. There is no standard approach to definingthe severity of esophageal lesions; our grading system wassimilar to that previously described.18 Finally, we did nothave a group with no ETM, and as such we are not able toevaluate whether either form of ETM used in this study wasprotective compared to no monitoring.

Conclusions

Intraluminal ETM is a commonly used instrument aimedto decrease risk of esophageal injury and development ofAE fistula formation. In our study, the use of a multi-sensorprobe compared to a single-sensor probe resulted in a higherdetection of esophageal temperature rises along with a higherincidence of esophageal injury.

References

1. Calkins H, Kuck KH, Cappato R, Brugada J, Camm AJ, Chen SA,Crijns HJ, Damiano RJ Jr, Davies DW, DiMarco J, Edgerton J, Ellen-bogen K, Ezekowitz MD, Haines DE, Haissaguerre M, Hindricks G,Iesaka Y, Jackman W, Jalife J, Jais P, Kalman J, Keane D, Kim YH,Kirchhof P, Klein G, Kottkamp H, Kumagai K, Lindsay BD, MansourM, Marchlinski FE, McCarthy PM, Mont JL, Morady F, Nademanee K,Nakagawa H, Natale A, Nattel S, Packer DL, Pappone C, PrystowskyE, Raviele A, Reddy V, Ruskin JN, Shemin RJ, Tsao HM, Wilber D;Heart rhythm society task force on catheter and surgical ablation ofatrial fibrillation: 2012 HRS/EHRA/ECAS expert consensus statementon catheter and surgical ablation of atrial fibrillation: Recommenda-tions for patient selection, procedural techniques, patient managementand follow-up, definitions, endpoints, and research trial design. HeartRhythm 2012;9:632-696.

2. Zellerhoff S, Ullerich H, Lenze F, Meister T, Wasmer K, Monnig G,Kobe J, Milberg P, Bittner A, Domschke W, Breithardt G, Eckardt L:Damage to the esophagus after atrial fibrillation ablation: Just the tipof the iceberb? High prevalence of mediastinal changes diagnosed byendosonography. Circ Arrhythm Electrophysiol 2010;3:155-159.

3. Lemola K, Sneider M, Desjardins B, Case I, Han J, Good E, TamirisaK, Tsemo A, Chugh A, Bogun F, Pelosi F Jr, Kazerooni E, MoradyF, Oral H: Computed tomographic analysis of the anatomy of the leftatrium and the esophagus: Implications for left atrial catheter ablation.Circulation 2004;110:3655-3660.

4. Leite LR, Santos SN, Maia H, Henz BD, Giuseppin F, Oliverira A,Zanatta AR, Peres AK, Novakoski C, Barreto JR, Vassalo F, d’AvilaA, Singh SM: Luminal esophageal temperature monitoring with a de-flectable esophageal temperature probe and intracardiac echocardio-graphy may reduce esophageal injury during atrial fibrillation abla-tion procedures: Results of a pilot study. Circ Arrhythm Electrophysiol2011;4:149-156.

5. Yamasaki H, Tada H, Sekiguchi Y, Igarashi M, Arimoto T, Machino T,Ozawa M, Naruse Y, Kuroki K, Tsuneoka H, Ito Y, Murakoshi N, KugaK, Hyodo I, Aonuma K: Prevalence and characteristics of asymptomaticexcessive transmural injury after radiofrequency catheter ablation ofatrial fibrillation. Heart Rhythm 2011;8:826-832.

6. Schmidt M, Nolker G, Marschang H, Gutleben KJ, Schibgilla V, RittgerH, Sinha AM, Ritscher G, Mayer D, Brachmann J, Marrouche NF: Inci-dence of oesophageal wall injury post-pulmonary vein atrum isolationfor treatment of patients with atrial fibrillation. Eurospace 2008;10:205-209.

7. Tilz RR, Chun KR, Metzner A, Burchard A, Wissner E, KoektuerkB, Konstantinidou M, Nuyens D, De Potter T, Neven K, FurnkranzA, Ouyang F, Schmidt B: Unexpected high incidence of esophagealinjury following pulmonary vein isolation using robotic navigation. JCardiovasc Electrophysiol 2010;21:853-858.

964 Journal of Cardiovascular Electrophysiology Vol. 24, No. 9, September 2013

8. Halm U, Gaspar T, Zachaus M, Sack S, Arya A, Piorkowski C, KniggeI, Hindricks G, Husser D: Thermal esophageal lesions after radiofre-quency catheter ablation of left atrial arrhythmias. Am J Gastroenterol2010;105:551-556.

9. Martinek M, Bencsik G, Aichinger J, Hassanein S, Schoefl R, KuchinkaP, Nesser HJ, Purerfellner H: Esophageal damage during radiofrequencyablation of atrial fibrillation: Impact of energy settings, lesion sets,and esophageal visualization. J Cardiovasc Electrophysiol 2009;20:726-733.

10. Singh SM, d’Avila A, Doshi SK, Brugge WR, Bedford RA, Mela T,Ruskin JN, Reddy VY: Esophageal injury and temperature monitor-ing during atrial fibrillation ablation. Circ Arrhythm Electrophysiol2008;1:162-168.

11. Scanavacca MI, D’Avila A, Parga J, Sosa E: Left atrial-esophagealfistula following radiofrequency catheter ablation. J Cardiovasc Elec-trophysiol 2004;15:960-962.

12. Dagres N, Hindricks G, Kottkamp H, Sommer P, Gaspar T, BodeK, Arya A, Husser D, Rallidis LS, Kremastinos DT, Piorkowski C:Complications of atrial fibrillation ablation in a high-volume center in1,000 procedures: Still cause for concern? J Cardiovasc Electrophysiol2009;20:1014-1019.

13. Ghia KK, Chugh A, Good E, Pelosi F, Jongnarangsin K, Bogun F,Morady F, Oral H: A nationwide survey on the prevalence of atrio-esophageal fistula after left atrial radiofrequency ablation. J Interv CardElectrophysiol 2009;24:33-36.

14. Cummings JE, Schweikert RA, Saliba WI, Burkhardt JD, BrachmannJ, Gunther J, Schibgilla V, Verma A, Dery M, Drago JL, KilicaslanF, Natale A: Assessment of temperature, proximity, and course of theesophagus during radiofrequency ablation within the left atrium. Circu-lation 2005;112:459-464.

15. Koruth JS, Reddy VY, Miller MA, Patel KK, Coffey JO, FischerA, Gomes JA, Dukkipati S, D’Avila A, Mittnacht A: Mechanicalesophageal displacement during catheter ablation for atrial fibrillation.J Cardiovasc Electrophysiol 2012;23:147-154.

16. Tsuchiya T, Ashikaga K, Nakagawa S, Hayashida K, Kugimiya H:Atrial fibrillation ablation with esophageal cooling with a cooled water-irrigated intraesophageal balloon: A pilot study. J Cardiovasc Electro-physiol 2007;18:145-150.

17. Good E, Oral H, Lemola K, Han J, Tamirisa K, Igic P, Elmouchi D,Tschopp D, Reich S, Chugh A, Bogun F, Pelosi F Jr, Morady F:Movement of the esophagus during left atrial catheter ablation for atrialfibrillation. J Am Coll Cardiol 2005;46:2107-2110.

18. Contreras-Valdes FM, Heist EK, Danik SB, Barrett CD, Blendea D,Brugge WR, Ptaszek L, Ruskin JN, Mansour M: Severity of esophagealinjury predicts time to healing after radiofrequency catheter ablation foratrial fibrillation. Heart Rhythm 2011;8:1862-1868.

19. Deneke T, Bunz K, Bastian A, Pasler M, Anders H, LehmannR, Meuser W, de Groot JR, Horlitz M, Haberkorn R, Mugge A,Shin DI: Utility of esophageal temperature monitoring during pul-monary vein isolation for atrial fibrillation using duty-cycled phasedradiofrequency ablation. J Cardiovasc Electrophysiol 2011;22:255-261.

20. Metzner A, Schmidt B, Fuernkranz A, Wissner E, Tilz RR, Koester I,Konstantinidou M, Rillig A, Yoshiga Y, Mathew S, Ouyang F, KuckKH: Esophageal temperature change and esophageal thermal lesionsafter pulmonary vein isolation using the novel endoscopic ablation sys-tem. Heart Rhythm 2011;8:815-820.

21. Cummings JE, Barrett CD, Litwak KN, DI Biase L, Chowdhury P, OhS, Ching CK, Saliba WI, Schweikert RA, Burkhardt JD, DE Marco S,Armaganijan L, Natale A: Esophageal luminal temperature measure-ment underestimates esophageal tissue temperature during radiofre-quency ablation within the canine left atrium: Comparison between8 mm tip and open irrigation catheters. J Cardiovasc Electrophysiol2008;19:641-644.

22. Redfearn DP, Trim GM, Skanes AC, Petrellis B, Krahn AD, Yee R,Klein GJ: Esophageal temperature monitoring during radiofrequencyablation of atrial fibrillation. J Cardiovasc Electrophysiol 2005;16:589-593.

23. Aupperle H, Doll N, Walther T, Kornherr P, Ullmann C, Schoon HA,Mohr FW: Ablation of atrial fibrillation and esophageal injury: Effectsof energy source and ablation technique. J Thorac Cardiovasc Surg2005;130:1549-1554.

24. Berjano EJ, Hornero F: What affects esophageal injury during radiofre-quency ablation of the left atrium? An engineering study based onfinite-element analysis. Physiol Meas 2005;26:837-848.

25. Kuwahara T, Takahashi A, Kobori A, Miyazaki S, Takahashi Y, TakeiA, Nozato T, Hikita H, Sato A, Aonuma K: Safe and effective ablationof atrial fibrillation: Importance of esophageal temperature monitoringto avoid periesophageal nerve injury as a complication of pulmonaryvein isolation. J Cardiovasc Electrophysiol 2009;20:1-6.

26. Yokoyama K, Nakagawa H, Seres KA, Jung E, Merino J, Zou Y, IkedaA, Pitha JV, Lazzara R, Jackman WM: Canine model of esophagealinjury and left atrial-esophageal fistula after pulmonary vein isolation.Circulation 2007;116(Suppl II):II-490.

27. Furnkranz A, Chun KR, Metzner A, Nuyens D, Schmidt B, BurchardA, Tilz R, Ouyang F, Kuck KH: Esophageal endoscopy results afterpulmonary vein isolation using the single big cryoballoon technique. JCardiovasc Electrophysiol 2010;21:869-874.

28. Rillig A, Meyerfeldt U, Birkemeyer R, Wiest S, Sauer BM, StaritzM, Jung W: Oesophageal temperature monitoring and incidence of oe-sophageal lesions after pulmonary vein isolation using a remote roboticnavigation system. Eurospace 2010;12:655-661.