Author + information
- Received June 3, 2018
- Revision received July 23, 2018
- Accepted July 26, 2018
- Published online December 17, 2018.
- Alex Baher, MDa,b,∗,
- Mobin Kheirkhahan, MDb,∗,
- Stephen J. Rechenmacher, MDa,
- Qussay Marashly, MBBSb,
- Eugene G. Kholmovski, PhDb,c,d,
- Johannes Siebermair, MD MHBAb,e,f,
- Madan Acharya, MDa,
- Mossab Aljuaid, MDa,
- Alan K. Morris, MSb,
- Gagandeep Kaur, BSb,
- Frederick T. Han, MDa,b,
- Brent D. Wilson, MD PhDa,b,
- Benjamin A. Steinberg, MD, MHSa,b,
- Nassir F. Marrouche, MDa,b and
- Mihail G. Chelu, MD, PhDa,b,∗ ()
- aDivision of Cardiovascular Medicine, University of Utah, Salt Lake City, Utah
- bComprehensive Arrhythmia Research & Management Center, University of Utah, Salt Lake City, Utah
- cUtah Center for Advanced Imaging Research, University of Utah, Salt Lake City, Utah
- dDepartment of Radiology and Imaging Sciences, University of Utah, Salt Lake City, Utah
- eDepartment of Medicine I, Grosshadern Clinic, University of Munich, Munich, Germany
- fGerman Cardiovascular Research Center partner site, Munich Heart Alliance, Munich, Germany
- ↵∗Address for correspondence:
Dr. Mihail G. Chelu, Division of Cardiology, University of Utah School of Medicine, 50 North Medical Drive, Salt Lake City, Utah 84123.
Objectives This study retrospectively evaluated the feasibility and esophageal thermal injury (ETI) patterns of high-power short-duration (HPSD) radiofrequency atrial fibrillation (AF) ablation.
Background ETI following AF ablation can lead to serious complications. Little consensus exists on the optimal radiofrequency power setting or on the optimal strategy to assess ETI.
Methods A total of 687 patients undergoing first-time AF ablation with either HPSD ablation (50 W for 5 s, n = 574) or low-power long-duration ablation (LPLD, ≤35 W for 10 to 30 s, n = 113) were analyzed. ETI was assessed by late gadolinium enhancement (LGE) magnetic resonance imaging (MRI) within 24 h post-ablation. Patients with moderate or severe esophageal LGE had a follow-up MRI within 24 h to 1 week, and esophagogastroduodenoscopies were performed when significant gastrointestinal symptoms or persistent LGE on repeat MRI was present. AF recurrence adjusted for potential confounders was analyzed.
Results The average age was 69.0 ± 11.8 years in the group undergoing HPSD ablation versus 68.3 ± 11.6 years in the LPLD group (p = 0.554), with 67.1% versus 59.3% male (p = 0.111). Esophageal LGE patterns were similar (64.8% vs. 57.5% none, 21.0% vs. 28.3% mild, 11.5% vs. 11.5% moderate, 2.8% vs. 2.7% severe for HPSD vs. LPLD, respectively; p = 0.370) with no atrioesophageal fistulas. Mean procedure length was significantly shorter in the HPSD group (149 ± 65 min vs. 251 ± 101 min; p < 0.001). AF recurrence rates were similar in the 2 groups for the mean 2.5-year follow-up period (adjusted, 42% vs. 41%; p = 0.571).
Conclusions HPSD ablation results in similar ETI patterns, as assessed by same-day LGE MRI, compared with the LPLD setting but with significantly shorter procedure times. Recurrence rates at 2.5-year follow-up are similar.
- atrial fibrillation
- esophageal injury
- magnetic resonance imaging
- power setting
- radiofrequency ablation
- thermal injury
Atrial fibrillation (AF) ablation is an effective treatment for symptomatic, drug-refractory AF (1). The approach to radiofrequency catheter ablation (RFCA) of AF involves circumferential isolation of pulmonary veins, as well as additional lesions to the left atrium in select patients. Although RFCA is generally considered safe, it still has inherent risks (2). Atrioesophageal fistula, 1 of the most dreaded complications of AF ablation, has a low incidence but can be fatal (3). In addition, other, less severe esophageal complications, such as ulceration and esophageal dysmotility caused by damage to the vagus nerve plexuses, are being increasingly recognized (4). Late gadolinium enhancement (LGE) magnetic resonance imaging (MRI) can be an effective tool for detection of early esophageal thermal injury (ETI) and for tracking its progression post-ablation (5). Most acute thermal injuries to the esophagus resolve within 1 week; however, persistent injury warrants further work-up (5,6).
Currently, there is no consensus on the best RFCA power setting for creating effective transmural lesions in the posterior left atrial (LA) wall while minimizing risk of ETI; however, a lower power (10 to 35 W) for a duration of 10 to 30 s is commonly used (7,8).
In this study, we report our experience using high-power short-duration (HPSD) and low-power long-duration (LPLD) irrigated-catheter energy for RFCA of AF and serial post-ablation LGE MRI examinations to assess the extent and progression of ablation-related acute ETI. The objective of this analysis was to assess procedural efficiency and acute post-ablation thermal injury patterns following AF ablation in HPSD and LPLD groups.
This was a retrospective study of consecutive patients who underwent RFCA ablation for AF at the University of Utah in Salt Lake City, Utah between 2007 and 2016. The study database was approved by the University of Utah Institutional Review Board.
Patients were included in the study if they were 18 years of age or older, had symptomatic AF, were referred for first-time AF RFCA ablation, and underwent RFCA using an open-tip irrigated catheter with either an HPSD or LPLD setting. Patients who did not have 24-h post-ablation LGE MRI and patients with unreported ablation power were excluded from the study (Figure 1). All patients were followed in clinic, and the presence of gastrointestinal (GI) symptoms post-ablation was assessed.
The LA ablation procedure was previously described in detail (9). In brief, double transseptal punctures were performed under intracardiac echocardiography guidance. A 10- or 20-pole circular mapping catheter and an irrigated ablation catheter were positioned in the left atrium. Electroanatomic mapping was performed using either CARTO (Biosense Webster, Diamond Bar, California) or ESI-NavX (St. Jude Medical, Minneapolis, Minnesota) systems.
All patients had pulmonary vein isolation, with additional linear ablations and/or substrate-based posterior wall debulking (on the basis of pre-ablation LA LGE) in select patients at the operator’s discretion. Linear ablations included the LA roof and mitral isthmus in patients with evidence of macro-re-entrant arrhythmias following pulmonary vein isolation.
Ablations were performed using either a non–contact force (non-CF) sensing catheter (Flexibility [St. Jude Medical] and ThermoCool NaviStar [Biosense Webster]), or a CF sensing catheter (TactiCath Quartz [St. Jude Medical], ThermoCool SmartTouch NaviStar [Biosense Webster], and ThermoCool SurroundFlow NaviStar [Biosense Webster]). Lesions were delivered using either the SmartAblate (Stockert, Freiburg, Germany) or Ampere (St. Jude Medical) radiofrequency generators. We used standard irrigation flow rates recommended by the catheter manufacturers. Contiguous ablations were made by dragging the catheter along the LA wall with a target interlesion distance of 4 to 5 mm.
The 2 ablation settings used were at the discretion of the individual operators: HPSD (50 W power for 5 s) and LPLD (≤35 W power for 10 to 30 s) with a maximum tip temperature of 50°C. In the HPSD group, power was kept constant throughout the procedure, whereas in the LPLD group, power was allowed to be decreased to 25 W when ablating the posterior wall at the discretion of the operator, but it never exceeded 35 W. A CF of 10 to 20 g was targeted (when using a CF catheter) or intracardiac echocardiography-guided visualization was used to ensure adequate catheter contact. An esophageal temperature probe was used in all patients, and ablation was stopped with increasing temperature of ≥2°C.
Short-term procedural success was defined as complete isolation of all pulmonary veins, bidirectional block across all linear ablations, and completion of fibrosis-based substrate modification.
MRI image acquisition and assessment of post-ablation ETI
MRI image acquisition was performed as previously described (9). Images were acquired with either 1.5-Tesla (T) Avanto or Aera or 3-T Verio clinical MRI scanners (Siemens Healthcare, Erlangen, Germany). Utah classification was assigned to patients on the basis of pre-ablation LA LGE (stage I, <10%; stage II, 10% to 20%; stage III, 20% to 30%; stage 4, >30%) (10).
Esophageal wall was manually segmented using Corview image processing software (Marrek, Inc., Salt Lake City, Utah). Segmentation was performed by tracing the inner and outer borders of esophagus. ETI was defined as any LGE in this area ascertained by manual thresholding by expert observers. Esophageal images displaying LGE were then segmented into 3-dimensional models, along with 3-dimensional segmentation and modeling of the LA wall, and were volume rendered to detect regions where enhancement overlapped. Esophageal enhancement was categorized as none (no detectable LGE), mild (minimal or focal LGE), moderate (transmural or nearly transmural LGE of the anterior wall), and severe (transmural LGE involving more than 5 mm of tissue or in more than 1 location) (5).
Post-operative care and follow-up
Following RFCA, all patients were observed for at least 24 h. Patients were continued on anticoagulation therapy perioperatively and post-operatively for a minimum of 2 months. GI symptoms were assessed post-operatively, as well as at the 3-month clinic follow-up. All patients were discharged on a 30-day course of an oral proton pump inhibitor.
LGE MRI–based esophageal monitoring
All patients underwent LGE MRI within 24 h post-ablation. Patients with severe acute esophageal enhancement had a follow-up LGE MRI within 24 to 48 h. Patients with moderate enhancement had a repeat scan within 1 week. If significant esophageal LGE persisted, repeat LGE MRI could not be scheduled or was contraindicated, or significant GI symptoms were reported, the patient was referred for an inpatient esophagogastroduodenoscopy (EGD) (5). All patients underwent repeat LGE MRIs before their 3-month follow-up visit (Figure 2).
AF recurrence follow-up
All patients received either 30-day or 60-day event monitors immediately following ablation and at the 3-month follow-up. The first 90 days post-ablation were considered the blanking period. Patients were followed at 3-, 6-, and 12-month intervals. If patients complained of AF symptoms, additional home monitoring was ordered. Recurrence was defined as sustained (≥30 s) AF or flutter on event monitors, AF or flutter on electrocardiogram, or the need for additional ablation and/or cardioversion (11).
Continuous variables were expressed as the mean ± SD. Student’s t-test was used to compare the continuous variables, and the Pearson chi-square test was used to compare the categorical variables. A multivariate, binary logistic regression analysis was used to assess independent predictors of moderate or severe esophageal enhancement on acute LGE MRI on the basis of all potentially confounding factors in the baseline characteristics. For arrhythmia recurrence, a survival analysis was performed, Kaplan-Meier curves were drawn, and significance was assessed with a log-rank test. The recurrence rates were adjusted for all baseline characteristics that were significantly different between the 2 groups. All statistical analyses were performed using R software version 3.4.0 open source software (R Foundation, Vienna, Austria). A 2-sided p value < 0.05 was considered statistically significant.
A total of 687 patients were included in the study (Figure 1); 574 of these patients (385 men; mean age 69.0 ± 11.8 years) belonged to the HPSD group, and 113 patients (67 men; mean age 68.3 ± 11.6 years) belonged to the LPLD group. A detailed list of baseline characteristics is shown in Table 1. Of these ablations, 146 (107 HPSD vs. 39 LPLD) were performed using a CF catheter. Further details of baseline characteristics of these patients are provided in the Online Appendix (Online Table 1).
Post-ablation ETI detected by LGE MRI
The incidence and severity of acute ETI detected on the immediate post-ablation LGE MRI were distributed similarly between the 2 groups (64.8% vs. 57.5% with none, 21.0% vs. 28.3% with mild, 11.5% vs. 11.5% with moderate, and 2.8% vs. 2.7%, with severe enhancement; p = 0.370), without evidence of atrioesophageal fistula in either group (Figure 3). There was no evidence of pre-ablation esophageal enhancement in any patients who had an available LGE MRI study (500 [87.1%] in the HPSD group vs. 91 [80.1%] in the LPLD group).
Moderate or severe (Figure 4A) acute post-ablation esophageal LGE was present in 82 (14.3%) HPSD-group patients and in 16 (14.3%) LPLD-group patients (p = 0.972). The esophageal enhancement either significantly improved or completely resolved in 42 (51.2%) of the patients in the HPSD group and in 7 (43.8%) of the patients in the LPLD group (p = 0.860) on repeat LGE MRI (Figure 4B, top). Patients with persistent esophageal enhancement (Figure 4B, bottom) on repeat LGE MRI (11 [13.4%] patients in the HPSD group and 1 [6.3%] patient in the LPLD group; p = 0.424) underwent inpatient EGDs. EGDs were also performed in all patients with severe esophageal enhancement who were not able to have a repeat LGE MRI study scheduled (4 in the HPSD group vs. 0 in the LPLD group). In some patients with moderate enhancement and no GI symptoms, a follow-up LGE MRI study was not performed, at the discretion of the operating physician (18 in the HPSD group vs. 4 in the LPLD group). In some patients with moderate esophageal enhancement, who had significant post-ablation GI symptoms, an EGD was performed directly without a follow-up LGE-MRI (7 in the HPSD group and 4 in LPLD group). Overall, EGDs were performed in 28 patients (23 in the HPSD group vs. 5 in the LPLD group). Figure 5 summarizes the results of the repeat LGE MRI and EGD studies. Esophageal enhancement subsided in most patients who had repeat LGE MRI. None of the patients with initial moderate esophageal enhancement that persisted on repeat LGE MRI had esophageal ulcers on EGD. Esophageal ulcers were detected in 1 of 3 patients who had persistent esophageal enhancement after initial severe enhancement. Esophageal ulcers improved on repeat EGD (Figures 6A to 6F). The 3-month LGE MRI study in all patients showed minimal to no esophageal enhancement.
Factors contributing to significant delayed esophageal enhancement
To assess factors that independently contributed to the presence of moderate or severe ETI on post-ablation LGE MRI, we performed a multivariate logistic regression. According to our analysis, the use of the CF catheter and ablation in the posterior wall was associated with 2.03 (p = 0.043) and 1.86 (p = 0.027) odds, respectively, of significant esophageal enhancement (see Online Appendix, Online Figures 1 to 3).
The procedure length (time from first access to last catheter removal) was shorter in the HPSD group compared with the LPLD group (149 ± 65 min vs 251 ± 101 min; p < 0.001). Ablation time was available in 429 (63.2%) patients (342 in the HSPS group and 87 in the LPLD group) and was also significantly shorter in the HPSD group (37.9 ± 13.9 mins vs. 55.0 ± 19.2 mins; p < 0.001) (Figures 7A and 7B).
AF recurrence at follow-up
The mean follow-up time of the study was 917 days. Figure 8 shows freedom from AF rates for the 2 groups. Overall, freedom from AF was similar (58% HPSD vs. 59% LPLD; p = 0.571) when adjusting for potential confounders including CHA2DS2-VASc (congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex) score, AF type, procedure type, and catheter type (Figure 7).
In this study, we described the feasibility and esophageal safety of HPSD irrigated-catheter AF ablation. Acute ETI post-ablation was assessed using LGE MRI, a novel and noninvasive imaging modality. Our study showed that the use of HPSD ablation: 1) results in similar ETI patterns on LGE MRI; 2) is associated with significantly shorter procedure times; and 3) produces similar long-term outcomes compared with LPLD ablation.
The use of 20-s to 30-s lesions at a low power setting comes from in vivo experiments using nonirrigated catheters. Simmers et al. (12) showed that the use of a 4-mm nonirrigated catheter in beagle ventricles by using 25 W for 30 s resulted in a 7.25-mm lesion depth. In the left atrium, where the mean posterior wall thickness is 1.5 to 2.5 mm (5,13,14) and the esophageal distance from the posterior left atrium is as small as 2.5 mm (15), such lesion sizes could result in thermal injury to the esophagus.
Bhaskaran et al. (16) recently showed that the use of 50 W for 5 s with an irrigated catheter resulted in in vitro and in vivo lesion depths of 2.2 ± 0.0 mm and 2.3 ± 0.5 mm, respectively. This is in comparison with lesion depths of 2.7 ± 0.1 mm and 2.4 ± 0.8 mm when using 40 W for 30 s. Both power settings resulted in transmural lesions in vivo, but there was a 10.5% incidence of a steam pops when using the 40-W for 30-s setting. These investigators also observed evidence of thermal lung injury with the 40-W setting, injury that was absent with the 50-W setting.
Winkle et al. (17) also reported the use of the HPSD setting in their AF ablations between 2003 and 2009. Although these investigators did not assess ETI in individual patients, they showed that the HPSD setting with an open-tip irrigated catheter resulted in few major complications and better long-term outcomes compared with the use of 40- and 45-W power settings with an open-tip irrigated catheter or a 70-W power setting with an 8-mm closed-tip catheter. However, their data predated routine use of most contemporary CF catheters, which are included in our data.
Our study compared the use of HPSD and LPLD settings in a large group of patients while assessing the extent of ETI in every patient by post-ablation LGE MRI. We previously showed that LGE MRI is a useful screening tool to assess ETI immediately post-ablation and that persistent esophageal enhancement on repeat scan is associated with a higher incidence of positive findings on EGD (5). Overall, we found moderate or severe esophageal enhancement in 14.2% of those patients. This finding is similar to the incidence of thermal injury detected on endoscopy in a previous study that evaluated the incidence of thermal injury in a large group of patients with post-ablation EGDs (18). In that study, Knopp et al. (18) retrospectively evaluated the incidence of esophageal findings in 425 patients who underwent post-ablation EGDs and found evidence of thermal lesions in 11% (n = 48) of these patients. LGE MRI is more advantageous than EGD as a screening tool because it is noninvasive. It also scans the entire thickness of the esophageal wall for thermal injury, whereas EGD can show only the luminal surface.
Our results show that HPSD and LPLD energy settings resulted in similar patterns of esophageal enhancement on same-day LGE MRIs, even though HPSD-group patients had significantly more ablation on the posterior wall, thus predisposing them to a higher risk of ETI. On the basis of our multivariate regression analysis, factors such as AF type, LA volume index, body mass index, and sex did not seem to correlate with significant esophageal enhancement, whereas posterior wall debulking and the use of CF catheters seemed to be associated with at least moderate esophageal enhancement. There have been recent reports of the potential increase of atrioesophageal fistulas with CF catheters (19). In our study, there were no atrioesophageal fistulas with CF catheters, and the distribution of severity of ETI was similar between the HPSD and LPLD groups. It is important to note that the risk of atrioesophageal fistulas can be minimized even with an HPSD setting as long as the ablation parameters are appropriately titrated on the posterior wall (i.e., short duration of 5 s or less, and reduced CF on the posterior wall to 10 to 15 g) (20).
Another major finding of our study is the significant reduction of procedure and ablation times in the HPSD group. This is important because longer ablation times have been associated with increased post-ablation cognitive dysfunction (21) and worsening heart failure in patients with left ventricular dysfunction as a result of increased fluid load from catheter irrigation (22).
HPSD ablation had similar efficacy compared with that of LPLD ablation. In our follow-up data, when adjusting for potential confounders in the baseline characteristics (CHA2DS2-VASc score, AF type, procedure type, and catheter type), the AF recurrence rates were similar. Although hypothesis generating, these data indicate that while significantly reducing the ablation time, HPSD ablation may not compromise procedure efficacy in terms of freedom from AF recurrence.
First, this was a single-center observational study and will need appropriate validation. Second, our study included only patients who had same-day post-ablation LGE –MRI studies, and therefore the outcomes may not reflect all patients who underwent RFCA of their AF with either of the settings. The major factors that precluded patients from having post-ablation LGE MRI were the presence of an implanted cardiac device or elevated serum creatinine levels. Third, choice of the HPSD versus the LPLD strategy was at the discretion of the operator and was not randomized. Fourth, the LPLD group in our cohort was smaller than the HPSD group and hence may have been more prone to any selection bias. Patients in the HSPD group also had significantly higher CHA2DS2-VASc scores and persistent AF. The HSPD group also had a more significant portion ablated with a non-CF catheter. We did adjust for these factors in our survival analysis, as well as provide a separate analysis for CF catheter–only patients in the Online Appendix. Finally, continuation of antiarrhythmic medications post-ablation was at the discretion of the operating physician and was not factored in our AF recurrence survival analysis. The primary objective of the study was to assess safety, feasibility, and efficiency of HPSD ablation. Future prospective studies to assess effectiveness of HPSD ablation in comparison with LPLD are welcomed.
AF ablation using HPSD, compared with LPLD, is feasible and results in similar ETI patterns detected on post-ablation LGE MRI. HSPD ablation results in significantly reduced procedure times.
COMPENTENCY IN MEDICAL KNOWLEDGE: ETI after AF ablation is a dreaded complication. Currently there is no consensus on optimal catheter energy setting during ablation or a robust noninvasive method to survey esophageal thermal damage post-ablation. Our study showed that using an HPSD ablation setting is safe for AF ablation and potentially saves time. We also showed that routine serial LGE MRI studies may be used as noninvasive tools to assess esophageal thermal damage.
TRANSLATIONAL OUTLOOK: Randomized controlled trials are needed to further assess the long-term AF recurrence outcomes using our proposed ablation energy setting.
↵∗ Drs. Baher and Kheirkhahan contributed equally to this paper and are joint first authors.
Dr. Kholmovski has equity interest in Marrek, Inc.; and has received consultant’s fees from Marrek, Inc. Dr. Morris has equity interest in Marrek, Inc. Dr. Han has received research grants from Boston Scientific and Abbott. Dr. Steinberg has received research support from Boston Scientific, St. Jude (Abbott), and Janssen; and has been a consultant for Biosense-Webster and Janssen. Dr. Marrouche has received consulting fees from Abbott, Biotronik, Wavelet Health, Cardiac Design, Medtronic, Preventice, Vytronus, Biosense Webster, Marrek, Inc., and Boston Scientific; has received research funding from Abbott, Boston Scientific, GE Healthcare, Siemens, Biotronik, Vytronus, and Biosense Webster; has an ownership interest in Marrek, Inc., and Cardiac Designs; and has conducted contracted research with Biosense Webster, Medtronic, St. Jude Medical, and Boston Scientific. Dr. Chelu has received research funding from Wavelet Health, Biotronik, Medtronic, and Boston Scientific. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- contact force
- congestive heart failure, hypertension, age, diabetes, stroke, vascular disease, sex
- esophageal thermal injury
- high-power short-duration
- left atrial
- late gadolinium enhancement
- low-power long-duration
- magnetic resonance imaging
- radiofrequency catheter ablation
- Received June 3, 2018.
- Revision received July 23, 2018.
- Accepted July 26, 2018.
- 2018 American College of Cardiology Foundation
- January C.T.,
- Wann L.S.,
- Alpert J.S.,
- et al.
- Cappato R.,
- Calkins H.,
- Chen S.A.,
- et al.
- Garg L.,
- Garg J.,
- Gupta N.,
- et al.
- Badger T.J.,
- Adjei-Poku Y.A.,
- Burgon N.S.,
- et al.
- Frankel D.S.
- Segerson N.M.,
- Daccarett M.,
- Badger T.J.,
- et al.
- Calkins H.,
- Kuck K.H.,
- Cappato R.,
- et al.
- Bhaskaran A.,
- Chik W.,
- Pouliopoulos J.,
- et al.
- Black-Maier E.,
- Pokorney S.D.,
- Barnett A.S.,
- et al.
- Chelu M.G.,
- Morris A.K.,
- Kholmovski E.G.,
- et al.
- Medi C.,
- Evered L.,
- Silbert B.,
- et al.