Esophageal Cooling For Protection During Left Atrial Ablation: A Systematic Review And Meta- 1 Analysis 2

30 Background: Thermal damage to the esophagus is a risk from radiofrequency (RF) ablation of the left 31 atrium for the treatment of atrial fibrillation (AF), with the most extreme type of thermal injury resulting 32 in atrio-esophageal fistula (AEF), with a correspondingly high mortality rate. Various approaches have 33 been developed to reduce esophageal injury, including power reduction, avoidance of greater contact- 34 force, esophageal deviation, and esophageal cooling. One method of esophageal cooling involves direct 35 instillation of cold water or saline into the esophagus during RF ablation. Although this method provides 36 limited heat-extraction capacity, studies of it have suggested potential benefit. Objective: We sought to perform a meta-analysis of existing studies evaluating esophageal cooling via 39 direct liquid instillation for the reduction of thermal injury. Methods: We reviewed Medline for existing studies involving esophageal cooling for protection of thermal injury during RF ablation. A meta-analysis was then performed using random effects model to calculate estimated effect size with 95% confidence intervals, with outcome of esophageal lesions, 44 stratified by severity, as determined by post-procedure endoscopy. Results: A total of 9 studies were identified and reviewed. After excluding pre-clinical and mathematical model studies, 3 were included in the meta-analysis, totaling 494 patients. Esophageal cooling showed a 48 tendency to shift lesion severity downward, such that total lesions did not show a statistically significant 49 change (OR 0.6, 95% CI 0.15 to 2.38). For high grade lesions, a significant OR of 0.39 (95% CI 0.17 to 50 0.89) in favor of esophageal cooling found, suggesting that esophageal cooling, even utilizing a low- capacity thermal extraction technique, reduces lesion severity from RF ablation.


Conclusions:
Esophageal cooling reduces lesion severity encountered during RF ablation, even when 54 using relatively low heat extraction methods such as direct instillation of cold liquid. Further 55 investigation of this approach is warranted. 56

INTRODUCTION 58
Thermal damage to the esophagus is a risk from radiofrequency (RF) ablation or cryoablation of the left 59 atrium for the treatment of atrial fibrillation (AF).[1-3] The most extreme type of thermal injury is an 60 atrio-esophageal fistula (AEF), with a mortality rate of 80% or more.

Data sources and search strategy 77
We utilized PubMed to perform a search of the literature from 1985 (prior to the earliest reports of 78 endocardial ablation to treat atrial fibrillation) to June 2019 for studies published on esophageal cooling 79 during cardiac ablation using a broad search with the following Boolean structure: (esophag* OR 80 oesophag*) AND cooling AND (ablation OR fibrillation) We did not restrict to English only. Details of 81 the systematic review were submitted for registration in PROSPERO on June 21, 2019. 82

Eligibility criteria 83
We excluded pre-clinical studies, bench-top, agar phantom, and mathematical model studies, and studies 84 that did not include formal endoscopy as an outcome measure. 85

Data collection 86
The primary data of interest were endoscopically identified lesions found after RF ablation. Because we 87 anticipated inconsistency in the categorization of lesion severity, we decided to simplify all lesion 88 severity measurement into severe lesions characterized by the presence of ulceration and mild to moderate 89 lesions encompassing all other abnormalities. 90

Statistical analysis 91
We input study data into Review Manager 5.3 to perform meta-analysis of the data entered, and present 92 the results graphically. SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for additional 93 analyses. The Cochran-Mantel-Haenszel (CMH) method was employed to test the null hypothesis that 94 the response rate is the same for the two arms (control versus treatment), after adjusting for possible 95 differences in study response rates. Furthermore, we fitted a random effect model using SAS procedures 96 GLIMMIX and NLMIXED by treating the study as a random effect. Because lesion grades are often 97 considered to be dichotomized into those that are likely to progress to atrio-esophageal fistula, and those 98 that are not, we initially analyzed the data as a binary outcome (high grade lesions concerning for 99 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint progression versus low grade lesions likely to heal spontaneously). Then, to further estimate effect size, 100 we used an ordinal logistic random intercept model, taking into account the ordered nature of lesion 101 grading (low to high, numerically). 102

RESULTS 103
A total of 9 studies were identified using the above criteria and additional search strategy. cooling did not decrease the overall incidence of thermal lesions, but noted a trend toward fewer severe 118 lesions with cooling (Fig 1). Esophageal thermal lesions were graded as follows: grade 0: no esophageal 119 lesion; grade 1: mucosal damage <1 cm width; grade 2: mucosal damage 1-3 cm width; grade 3: mucosal 120 damage > 3 cm width or visualization of deeper layer; and grade 4: bleeding ulcer or with overlying clot. 121 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint [18] The severity of the esophageal lesions was qualitatively graded as mild, moderate, or severe, 127 according to their extent and color, with the severe category corresponding to those categorized as grade 128 III or IV by John et al. The authors found that this approach reduced the severity of esophageal lesions, 129 but did not reduce their incidence: lesions occurred in 20% of the experimental group, and 22% of the 130 controls, with 3 moderate and 7 mild in the cooled group and 3 severe, 1 moderate, and 7 mild in the 131 control (Fig 2). 132 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint The percentage of patients free from any ulceration or erosion in each group was found to be 63.6%, 142 87.5%, and 95.2%, respectively (Fig 3). Esophageal lesions were classified as normal (score 1), erosion 143 (patchy mucosal ulceration: score 2), mild ulcer (necrosis less than 3 mm in diameter with red spot: score 144 3), severe ulcer (necrosis more than 3 mm in diameter with red spot and/or with a hemorrhagic 145 appearance, often with fibrinoid material: score 4    Separately, utilizing the CMH method, we obtained a significant p-value of 0.016 indicating the 173 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint association between treatment and grade remains strong. Furthermore, in a binary logistic regression 174 model, an OR of 0.46 (95% CI 0.28 to 0.75) was found. 175 heat extraction capacity of this approach, our meta-analysis of 3 studies suggests that cooling in this 192 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint manner offers a clinically significant protective effect from severe lesions, and provides a 61% reduction 193 in high grade lesion formation (with a 95% CI of 11% to 83% reduction). In contrast, our study had the benefit of an additional publication, and we analyzed data stratified by 212 severity of lesions. Although a detailed understanding of the mechanisms of AEF formation is still 213 developing, there is general agreement that thermal injury is a precursor, and that higher-grade thermal 214 injury has higher risk of progression to AEF.
[24] 215 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint Taking into account each lesion grade independently in a statistical model allows an alternative approach 216 to estimate effect size that may provide further refinement of the estimate, at the cost of decreased 217 precision. We found a point estimate using this approach suggesting a reduction in high-grade lesions of -218 23%, although a higher number of patients with a greater number of high-grade lesions are necessary to 219 narrow the confidence intervals around this estimate, which with the population included here, ranges 220 from -85% to +38%. On the other hand, utilizing a methodology with greater performance and higher 221 heat extraction capacity may increase the effect size point estimate further when either binary logistic or 222 ordinal regression is utilized. 223

LIMITATIONS 225
The studies analyzed in this report differed in specific RF techniques and equipment. Nevertheless, all 226 used radiofrequency ablation, and the variation in technology reflects real-world practice currently. The 227 studies reviewed did not all utilize randomization, and there was no description of any attempt at blinding 228 the patients to the protection strategy used. Lesion grading varied between studies, but each scale 229 involved a lesion characterization that allowed stratifying on common ground. Patient characteristics and 230 ablation technologies and techniques utilized in the evaluated studies differed in many aspects; however, 231 this may serve to broaden the generalizability of these findings. 232

CONCLUSIONS 234
Esophageal cooling reduces lesion severity encountered during RF ablation, even when using relatively 235 low heat extraction methods such as direct instillation of small amounts of cold liquid. Further 236 investigation of this approach is warranted, particularly utilizing higher heat extraction technology such as 237 closed-circuit devices with high coolant flow rates. 238 . CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint  CC-BY-NC-ND 4.0 International license It is made available under a author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/19003228 doi: medRxiv preprint