Epicardial adipose tissue affects the efficacy of left atrial posterior wall isolation for persistent atrial fibrillation

Abstract Background Epicardial adipose tissue (EAT) contributes to atrial fibrillation (AF). However, its impact on the efficacy of left atrial posterior wall isolation (LAPWI) is unclear. Methods Forty‐four nonparoxysmal AF patients underwent LAPWI after pulmonary vein isolation. EAT overlap on LAPWI was assessed by fusing computed tomography images with electro‐anatomical mapping. Results During the 21 ± 7 months of follow‐up, AF recurred in 10 patients (23%). The total and left atrial EAT volumes were 113 ± 36 and 33 ± 12 cm3, respectively. No differences were found between the AF‐free and AF‐recurrent groups regarding EAT volume. The EAT overlaps on LAPWI lines and LAPWI area were 1.2 ± 1.0 and 0.5 ± 0.9 cm2 respectively. Although no difference was found between groups regarding the EAT overlap on LAPWI area, the AF‐free group had a significantly larger EAT overlap on LAPWI lines (1.4 ± 1.0 vs 0.6 ± 0.6 cm2, P = .014). Multivariate analysis identified EAT overlap on LAPWI lines as an independent predictor of AF recurrence (hazard ratio: 0.399, 95% confidence interval: 0.178‐0.891, P = .025). Kaplan‐Meier analysis revealed that, during follow‐up, 92% of the large EAT overlap group (≥1.0 cm2) and 58% of the small EAT overlap group (<1.0 cm2) remained AF‐free (P = .008). Conclusions EAT overlap on LAPWI lines is related to a high AF freedom rate. Direct radiofrequency application to EAT overlap may be necessary to suppress AF.


| INTRODUC TI ON
Epicardial adipose tissue (EAT) is a metabolically active tissue located between the visceral pericardium and the epicardium, known as a risk factor for various cardiovascular diseases, 1 particularly atrial fibrillation (AF), to which EAT is closely related.
EAT volume is related to both new-onset AF 2,3 and AF recurrence after catheter ablation 4,5 or cardioversion. 6 Moreover, EAT alters the atrial electrophysiological properties contributing to generate AF. 7,8 EAT promotes AF both through local and remote effects. The local effect involves adipocyte infiltration into the atrial myocardium 9,10 and fibrosis of the neighboring atrial tissue by paracrine secretion of profibrotic adipokines. [11][12][13] The remote effect, on the other hand, arises from the systemic secretion of adipokines and metabolites. 14 Furthermore, ganglionated plexuses are located in EAT, altering the electrophysiological properties of atrial remote areas. 15,16 It remains unclear whether the main pathway underlying the contribution of EAT for the development of AF involves the local or remote effect.
Previous studies 17,18 have revealed that applying radiofrequency to the areas overlapped with EAT suppressed AF recurrence after catheter ablation. However, if the local effect is the main pathway for the AF-promoting effect of EAT, isolation of the areas overlapping with EAT should be sufficient to suppress AF.
If, on the other hand, the remote effect contributes to AF development, applying radiofrequency directly to the areas overlapped with EAT should be needed. Since left atrial posterior wall isolation (LAPWI) creates an isolated area in the left atrium, assessing EAT overlap on LAPWI may provide new insights into the mechanisms of the AF-promoting effect of EAT. Therefore, the present study aimed to investigate the impact of EAT overlap on AF recurrence after catheter ablation, including LAPWI in patients with persistent AF.

| Study population
The present study retrospectively analyzed consecutive patients with persistent and long-standing persistent AF who underwent catheter ablation at the Toyama University Hospital (Toyama, Japan)

| Catheter ablation and mapping
All antiarrhythmic drugs were discontinued for at least five half-lives, and no patient received oral amiodarone prior to ablation. Sheath introducers were inserted through the right femoral vein under sedation. A transseptal procedure was performed, and two 8-F SL0 were performed in 342 ± 288 points. Fractionation analyses were performed on the NavX system for CFAE mapping. CFAE mapping was performed with the following settings: peak-to-peak sensitivity, 0.04 mV; electrogram refractory period, 30 ms; electrogram width, <10 ms; electrogram segment length, 5 seconds. 17 The fractionated index was defined as the average time interval between consecutive deflections over each recording period. CFAE was defined as an average fractionated index of ≤120 ms 19 The voltage map was constructed from the voltage data obtained during contact mapping. Low voltage areas were delineated based on a bipolar voltage of <0.5 mV (determined as the maximum bipolar voltage of three consecutive AF beats, excluding QRS complexes from the window of interest).
Following contact mapping, LAPWI was performed by creating a left atrial roof line at the most cranial aspect and a floor line joining the most inferior margin of the PVI line. If the left atrial posterior wall was not isolated, additional radiofrequency applications were performed, targeting the earliest activation site on the isolation lines. The entrance block was confirmed by voltage mapping using a 3D mapping system. The exit block was confirmed after external cardioversion using high output pacing within the LAPWI lines.
The procedure was completed with the creation of a block line on the mitral isthmus and the cavotricuspid isthmus. The low voltage areas and CFAE areas were not ablated. Each radiofrequency application was performed for 30-50 seconds. The temperature and power were maintained at 42°C and 30 W, respectively; a maximum power of 25 W was used while delivering energy to the sites near the esophagus.

| Multidetector enhanced computed tomography imaging
All patients underwent contrast-enhanced multidetector computed tomography (CT), which was performed using either a dual source 128-slice multidetector CT-scanner (Somatom Definition AS+; Siemens Medical Solutions, Forchheim, Germany; 0.30 seconds gantry rotation time, 120 kV) or a 64-slice multidetector CT-scanner (Somatom Sensation Cardiac 64; Siemens Medical Solutions; 0.33 seconds gantry rotation time, 120 kV) within 3 days prior to ablation. Images were captured during a breath hold at full expiration using cardiac gating from the aortic arch's caudal aspect, through the cranial aspect of the left hemidiaphragm. The slice was 0.6-mm thick, and a reconstructed series in systole (20%-40% of the R-R interval) was selected for analysis.

| Postprocedure care and follow-up
All patients were subjected to a clinical interview and a surface 12lead electrocardiogram on the day after ablation and thereafter during the monthly visits to the outpatient clinic. Twenty-four-hour Holter monitoring was performed on the day following catheter ablation and as needed thereafter the follow-up period. AF recurrence was identified both from the symptoms along with documentation of an AF episode lasting ≥30 seconds, on a surface 12-lead electrocardiogram or Holter monitoring after a 3-month blanking period from ablation. Antiarrhythmic drugs were resumed at the responsible physician's discretion.

| Statistical analysis
Results are presented as the mean ± standard deviation and 95% confidence intervals. Continuous variables were compared using the unpaired Student's t-test, whereas categorical variables were compared using the χ 2 test. Univariate and multivariate Cox proportional hazards regression analyses were performed to identify AF recurrence predictors. Related factors with a P < .100 in the univariate analysis were selected as independent variables for multivariate analysis. Upon observing a significant correlation between two variables, the variable with lower significance was excluded from the multivariate analysis to eliminate the multicollinearity. Correlations F I G U R E 1 Representative cases of epicardial adipose tissue (EAT) overlap on ablation lines and the left atrial posterior wall isolation (LAPWI) area. Fusion of the computed tomographic image with EAT (red structures) and the geometry created on three-dimensional mapping was performed. The red and green tags show the ablation points for pulmonary vein isolation (PVI) and LAPWI respectively. (A) Atrial fibrillation (AF)-free case. The EAT overlap on LAPWI lines and LAPWI area were 3.6 and 2.2 cm 2 respectively. (B) AF-recurrent case. The EAT overlap on LAPWI lines and LAPWI area were both 0 cm 2 between parameters were analyzed using the Pearson's correlation coefficient. Receiver operating characteristic curve analyses were performed to determine the optimal parameter cut-off values for AF recurrence prediction. The outcome of ablation over time was plotted using a Kaplan-Meier survival curve and compared between parameter groups (below vs above cut-off) by log-rank test.
A P < .050 (two-tailed) was accepted as statistically significant for all tests.

| Patients' characteristics and catheter ablation
Patients' characteristics are shown in Table 1. The mean age was 64 ± 10 years and 84% of the patients were male. The mean body mass index was 25 ± 3 kg/m 2 . The mean AF duration was 16 ± 21 months and 41% of the patients had long-standing persistent AF. Structural heart disease and congestive heart failure were observed in 32% and 20% of patients respectively. Prior to ablation, antiarrhythmic drugs were administered in only 11% of the patients.
The left atrium was enlarged, and the left atrial appendage flow velocity was below the normal range. The left ventricular ejection function was preserved and the total and left atrial EAT volumes were 113 ± 36 and 33 ± 12 cm 3 respectively.
LAPWI was successfully performed in 44 patients (100%). Mitral isthmus block and cavotricuspid isthmus block were achieved in 40 patients (91%) and 44 patients (100%) respectively. Catheter ablation data are shown in Table 2. The left atrial surface area measured on 3D mapping was 106 ± 18 cm 2 , and the LAPWI area was 11 ± 4 cm 2 . The total radiofrequency time was 66 ± 16 minutes, and the radiofrequency application time needed for LAPWI was 8 ± 5 minutes. The low voltage areas observed in the left atrium and posterior wall were 57 ± 21 and 10 ± 5 cm 2 respectively. The CFAE areas observed in the left atrium and posterior wall were 31 ± 17 and 2 ± 3 cm 2 respectively.
EAT overlap data are shown in Table 3

| Outcome of catheter ablation and EAT
AF recurred in 10 patients (23%) during 21 ± 7 months of follow-up.
Accordingly, the study subjects were divided into an AF-free group (n = 34 patients) and an AF-recurrent group (n = 10 patients).
No significant differences were found between groups regarding the patients' characteristics, including the rate of administration of antiarrhythmic drugs after catheter ablation, EAT volume (Table 1).
Regarding catheter ablation data, the left atrial surface area and LAPWI area were not different between groups (Table 2). Moreover, although the total radiofrequency time tended to be shorter in the AF-free group, no difference was found between groups regarding the radiofrequency time for LAPWI. Furthermore, no differences between groups were found in the low voltage and CFAE areas.
No significant differences were found between groups regarding EAT overlaps on PVI lines, floor line, and mitral isthmus line (Table 3); however, EAT overlap on the roof line was significantly larger in the AF-free group than in the AF-recurrent group (1.2 ± 1.0 vs 0.5 ± 0.5 cm 2 , P = .029). Moreover, although the EAT overlap on LAPWI area was not different between groups, the EAT overlap on LAPWI lines was significantly larger in the AF-free group than in the AF-recurrent group (1.4 ± 1.0 vs 0.6 ± 0.6 cm 2 , P = .014).
Univariate Cox regression analysis identified the left ventricular ejection fraction, total radiofrequency application time, and EAT

TA B L E 3 Epicardial adipose tissue (EAT) overlaps on ablation lines and the left atrial posterior wall isolation (LAPWI) area
overlap on LAPWI lines as factors for multivariate analysis (Table 4).
In the receiver operating characteristic curve analysis, the EAT

| D ISCUSS I ON
The present study evaluated the impact of EAT on AF recurrence after catheter ablation with LAPWI for persistent AF patients. The study's main findings were as follows: (a) EAT volume was not associated with AF recurrence after catheter ablation; (b) EAT overlap on the LAPWI area did not correlate with low voltage area or with CFAE area on the posterior wall; (c) Although EAT overlap on the LAPWI area was not related to AF recurrence, large EAT overlap on LAPWI ablation lines was associated with a high AF freedom rate after catheter ablation.
The local effect of EAT is partly demonstrated by the direct adipocyte infiltration into the underlying atrial myocardium. A previous experimental study using an ovine model revealed that obesity caused EAT accumulation with pronounced myocardial infiltration by adipocytes, particularly over the left atrial posterior wall. 9 Such direct fatty infiltration separating myocytes could lead to conduction slowing and heterogeneity. 20,21 Moreover, the fact that profibrotic adipokines secreted from EAT facilitate paracrine effects on the atrial myocardium is another possible mechanism of the local effect. A previous study 12  EAT also exerts a remote effect on the atrial myocardium, which arises from the systemic secretion of adipokines and metabolites.
For instance, increased leptin levels can lead to increased aldosterone secretion, endothelial dysfunction, increased vascular stiffness, hypertension, and cardiac hypertrophy, all of which could contribute for AF pathogenesis. 22 Moreover, adipokines and metabolites from EAT are associated with systemic inflammation, oxidative stress, and autonomic dysfunction, resulting in a pro-arrhythmogenic state. 14 Another potential mechanism underlying the remote effect involves the fact that EAT contains ganglion plexuses, which play a critical role for AF initiation and perpetuation. 23 Increased ganglion plexus activity induces AF through shortening action potential duration via parasympathetic stimulation and an increased calcium loading via sympathetic stimulation. 23 In the present study, direct application of radiofrequency to the area overlapped with EAT may have suppressed AF recurrence through modification of ganglion plexuses. This hypothesis is supported by previous studies that revealed the AF-suppressing effect of radiofrequency application targeting ganglion plexuses. 23,24 Furthermore, the finding of the present study that EAT overlap on the mitral isthmus line was not associated with AF recurrence is consistent with this hypothesis.
Because the major ganglion plexuses are located around the pulmonary vein antrum and the left atrial posterior wall. 23,24 The findings of the present study suggest that the AF-promoting effect of EAT is mediated mainly by the remote effect. If the local effect comprised the main pathway underlying EAT promoting AF recurrence, then EAT overlap on the LAPWI area should be related to AF recurrence after ablation. Previous studies reported the effectiveness of catheter ablation to target areas overlapped with EAT. 17,18 Applying radiofrequency to the high dominant frequency sites overlapped with EAT suppresses AF recurrence after catheter ablation. 17 Moreover, EAT-based ablation achieved relatively high AF freedom rate compared with generalized stepwise ablation. 18 Although the effectiveness of the direct radiofrequency application to the areas overlapped with EAT was revealed in these studies, 17,18 the effectiveness of isolating EAT overlapped area remains unclear.
Therefore, the present study is the first to suggest the remote effect as a primary cause of the AF-promoting effect of EAT.

| Clinical implications
The  Therefore, the detailed mechanism of the AF-suppressing effect of radiofrequency application to EAT overlap remains unclear. Further studies are required to assess the relationship between changes in the serum adipokine levels and EAT overlap on ablation sites.

| CON CLUS ION
Although EAT overlap on LAPWI did not relate to AF recurrence, large EAT overlap on the LAPWI ablation lines was associated with high AF freedom rate after catheter ablation. This suggests that the AF-promoting effect of EAT is more mediated by the remote effect than by the local effect. Therefore, direct radiofrequency application to EAT overlap may be necessary for AF suppression.

ACK N OWLED G M ENTS
The authors thank Mr Yasushi Terada and Mr Norihiko Konishi for their technical assistance.

CO N FLI C T O F I NTE R E S T
Authors declare no Conflict of Interests for this article.