Download Does Location of Epicardial Adipose Tissue Correspond to

Does Location of Epicardial Adipose Tissue Correspond to
Endocardial High Dominant Frequency or
Complex Fractionated Atrial Electrogram
Sites During Atrial Fibrillation?
Koichi Nagashima, MD; Yasuo Okumura, MD; Ichiro Watanabe, MD; Toshiko Nakai, MD;
Kimie Ohkubo, MD; Masayoshi Kofune, MD; Hiroaki Mano, MD; Kazumasa Sonoda, MD;
Takafumi Hiro, MD; Mizuki Nikaido, BE; Atsushi Hirayama, MD
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
Background—Although increased epicardial adipose tissue (EAT) volume is known to be associated with increased
prevalence of atrial fibrillation (AF), the exact mechanisms are unclear. Therefore, we investigated whether EAT locations
were associated with high dominant frequency (DF) sites or complicated fractionated atrial electrogram sites during AF.
Methods and Results—Three-dimensional reconstruction computed tomography images depicting EAT volumes (obtained
by 320-detector-row multislice computed tomography) were merged with NavX-based DF and complicated fractionated
atrial electrogram maps obtained during AF for 16 patients with paroxysmal AF and for 18 patients with persistent AF.
Agreement between locations of the EAT, especially EAT surrounding the left atrium, and of high DF or complicated
fractionated atrial electrogram sites was quantified. In addition, serum biomarker levels were determined. EAT surrounding
the left atrium volumes was significantly greater in patients with persistent AF than in patients with paroxysmal AF
(52.9 cm3 [95% CI, 44.2–61.5] versus 34.8 cm3 [95% CI, 26.6–43.0]; P=0.007). Serum high-sensitivity C-reactive protein
and interleukin-6 levels were significantly higher in persistent AF patients than in paroxysmal AF patients (median highsensitivity C-reactive protein, 969 ng/mL [interquartile range, 307–1678] versus 320 ng/mL [interquartile range, 120–
660]; P=0.008; median interleukin-6, 2.4 pg/mL [interquartile range, 1.7–3.2] versus 1.3 [interquartile range, 0.8–2.4]
pg/mL; P=0.017). EAT locations were in excellent agreement with high DF sites (κ=0.77 [95% CI, 0.71–0.82]) but in
poor agreement with complicated fractionated atrial electrogram sites (κ=0.22 [95% CI, 0.13–0.31]).
Conclusions—Increased EAT volume and elevation of inflammatory biomarkers are noted in persistent AF rather than
paroxysmal AF patients. High DF sites are located adjacent to EAT sites. Thus, EAT may be involved in the maintenance
of AF. (Circ Arrhythm Electrophysiol. 2012;5:676-683.)
Key Words: atrial fibrillation ◼ dominant frequency
◼ complex fractionated atrial electrogram ◼ epicardial adipose tissue
O
ver the past decade, catheter-based pulmonary vein isolation (PVI) has become a widely accepted therapy for
patients with symptomatic drug-refractory paroxysmal atrial
fibrillation (PAF).1 Termination of persistent AF (PerAF), however, often requires extensive ablation, including ablation at
complex fractionated atrial electrogram (CFAE) sites and high
dominant frequency (DF) sites and multiple linear ablation in
addition to PVI.2–4 There have been numerous reports regarding the role of CFAEs and high DFs in the maintenance of AF.
CFAEs are now considered to be a result of dyssynchronous activation of separate cell groups at pivot points or of wave collision,
far-field potentials, or repetitive activations of the AF driver(s) or
local reentry circuit,5–7 whereas high DF is reported to be related
to the center of a focal-firing rotor or local reentry circuit.8
Editorial see p 618
Clinical Perspective on p 683
Association of AF with metabolic syndrome and
inflammation is well established.9,10 Recent studies have
shown that increased epicardial adipose tissue (EAT) volume
is associated with increased prevalence of AF and that EAT
volume can predict the development of AF.11–14 EAT is known
to secrete several activated proinflammatory cytokines, such
as tumor necrosis factor-α, transforming growth factor-β,
and interleukin-6 (IL-6).15,16 Nevertheless, the exact role of
EAT in the maintenance of AF is unclear. Hypothesizing that
delineation of the distribution of EAT around the left atrium
(LA), PVs, and CFAE and high DF sites would provide insight
Received January 31, 2012; accepted June 4, 2012.
From the Division of Cardiology, Department of Medicine, Nihon University School of Medicine (K.N., Y.O., I.W., T.N., K.O., M.K., H.M., K.S., T.H.,
A.H.); and Nihon Kohden Co, Ltd (M.N.), Tokyo, Japan.
Correspondence to Yasuo Okumura, MD, Division of Cardiology, Department of Medicine, Nihon University School of Medicine, 30-1 Ohyaguchikamicho, Itabashi-ku, Tokyo 173–8610, Japan. E-mail [email protected]
© 2012 American Heart Association, Inc.
Circ Arrhythm Electrophysiol is available at http://circep.ahajournals.org
676
DOI: 10.1161/CIRCEP.112.971200
 Nagashima et al Epicardial Adipose Tissue and DF 677
into the mechanism responsible for maintenance of AF,
we investigated whether EAT locations correspond to high DF
or CFAE sites during AF.
Methods
Study Patients
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
The study involved 34 consecutive patients (29 men, 5 women;
mean age, 56±13 years) scheduled for first catheter ablation of AF.
Sixteen of these patients had PAF (ie, AF lasting <7 days), and 18
had PerAF (ie, AF lasting >7 days). No patients with congestive heart
failure or a history of ischemic heart disease, cardiomyopathy, valvular heart disease, or congenital heart disease were included. All
patients provided written informed consent for electrophysiological
study and ablation procedures. Adequate oral anticoagulation therapy
was given for at least 1 month before the ablation procedure, and
all antiarrhythmic drugs were discontinued for at least 5 half-lives
before the procedure. On admission, a medical history was obtained,
and physical examination, 12-lead electrocardiography, chest x-ray,
and transesophageal and transthoracic echocardiography were performed; values obtained were considered baseline values. The study
was approved by the Institutional Review Board of Nihon University
Itabashi Hospital, and all patients provided written informed consent
for their participation.
Hematologic Measurements
Blood samples were obtained from a femoral vein before ablation
in the electrophysiological laboratory. Serum hemoglobin A1c, total
cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, and triglyceride levels were measured. Specific
biomarkers related to AF17–20 were measured: serum high-sensitivity
C-reactive protein (hs-CRP) by particle-enhanced immunonephelometry (Behring Nephelometer II; Siemens Healthcare Diagnostics,
Eschborn, Germany), serum IL-6 by ELISA (with a commercially
available kit from Fujirebio Inc, Tokyo, Japan), and serum matrix
metalloproteinase-2 (MMP-2) by 1-step sandwich enzyme immunoassay (with a commercially available kit from Daiichi Fine Chemical
Co, Ltd, Toyama, Japan).
Echocardiographic Evaluation
In all patients, transthoracic echocardiography was performed 1
day before ablation with an ACUSON Sequoia C256 echocardiography system (Siemens Medical Solutions USA, Inc, Malvern, PA).
Maximum LA volume was calculated by the prolate ellipsoid method,21 and left ventricular ejection fraction was assessed in the parasternal long-axis view by the Teichholz method. Measurements from
3 consecutive beats were averaged.
Electrophysiological Study
Electrophysiological study was performed with patients under conscious sedation achieved with propofol and fentanyl. After vascular
access was obtained, single transseptal puncture was performed, and
intravenous heparin was administered to maintain an activated clotting time >300 seconds. After 3 long sheaths (2 SL0 sheaths and 1
SL1 sheath; St. Jude Medical Inc, St. Paul, MN) were inserted into
the LA via transseptal puncture, 3-dimensional (3D) geometry of the
LA and the 4 PVs was reconstructed with an EnSite NavX mapping
system (St. Jude Medical Inc, Minneapolis, MN) from data obtained
with a 20-pole circular mapping catheter (1.5-mm interelectrode distance; Livewire Spiral HP catheter; St. Jude Medical).
CFAE and Fast-Fourier Transform Analyses
For CFAE and fast-Fourier transform analyses, bipolar signals from
the mapping catheter were acquired during AF and filtered between
30 and 500 Hz. If patients were in sinus rhythm, AF was induced
by decremental atrial pacing, and analysis was done after a 5-minute stabilization period of the AF rhythm. Bipolar signals were recorded from the 4 PVs; the antra of each PV; the anterior, posterior,
and septal surfaces and roof and floor of the LA; the mitral annulus;
and the LA appendage (LAA). The NavX mapping parameters were
set to CFE-mean, by which an interval-analysis algorithm is used
to measure the average index of the fractionation at each site, and a
color map of the fractionation interval (CFAE map) was constructed.
The fractionation interval was defined as the average time interval
between consecutive deflections during a 5-second recording period.
The settings included a refractory period of 40 ms, a peak-to-peak
sensitivity between 0.05 mV and 0.1 mV, and a duration of 10 ms.
Continuous CFAEs were defined as having a mean fractionation
interval <50 ms and variable CFAEs as having a fractionation interval of 50 to 120 ms.22–24 For fast-Fourier transform analysis, DF
Figure 1. Twenty-two segments of the
left atrium and pulmonary veins (PVs)
for quantitative assessment of the distribution of epicardial adipose tissue,
high-dominant frequency sites, and
complex fractionated atrial electrograms.
LSPV indicates left superior PV; LIPV,
left inferior PV; RSPV, right superior PV;
RIPV, right inferior PV; LAA, left atrial
appendage; MV, mitral valve; AP, anteroposterior; PA, posteroanterior; RAO,
right anterior oblique; LAO, left anterior
oblique.
678 Circ Arrhythm Electrophysiol August 2012
distribution was analyzed, and a DF map was constructed by means
of DF software installed in the NavX mapping system (sampling rate
1200 Hz, resolution 0.14 Hz, with a Hamming window function).22,23
The bipolar signals obtained from the 5-second recording were analyzed, and the highest peak frequency of the resulting spectrum was
identified as the DF. On the DF map produced by the NavX system, a
high DF site was defined as a site with a frequency >8 Hz and colored
bright purple.25
Multidetector Computed Tomography and
EAT Measurements
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
Details of EAT detection have been described previously.11 In brief,
EAT volume was calculated from noncontrast images obtained with a
3D spiral computed tomography (CT) scanner (320-row detector, dynamic volume CT scanner; Aquilion ONE, Toshiba Medical Systems,
Tokyo, Japan; 0.35-second gantry rotation time, 120 kV, and 350–450
mA) within 1 week before PVI. Gated studies were performed under
an ECG-triggered scanning protocol. To minimize motion artifacts,
patients were given β-blockers and underwent CT scanning only if
their heart rate was >80 beats per minute. On a workstation (ZIO
M900 QUADRA; Amin Co, Ltd, Tokyo, Japan), total EAT was detected by assigning Hounsfield units ranging from −50 to −200 to
fat, and the total EAT volume was semiautomatically reconstructed
from contiguous 0.5-mm slices of axial images from the bifurcation
of the pulmonary artery to the diaphragm. Thereafter, the volume of
EAT surrounding the LA (LA-EAT) was manually segmented from
the total EAT, ie, it was obtained by deleting EAT volume from the
left ventricular side anterior to the mitral annulus and the right atrial
side anterior to the right superior PV and then from the lower side of
the coronary sinus from the total EAT, leaving the EAT surrounding
the LA.
In addition, axial CT images were transferred to the NavX mapping system equipped with NavX system image integration software
(EnSite Verismo; St. Jude Medical).26 The surface reconstruction of
LA plus PV volume was segmented from each chamber of interest.
LA-EAT was also segmented and reconstructed after detection by assignment of Hounsfield units from −50 to −200.
Quantitative Assessment of Distributions of EAT,
High DF Sites, and CFAEs
3D LA and LA-EAT CT images were merged with NavX-based DF
and CFAEs maps during AF, as previously reported.26 Thereafter, we
divided the PVs and LA into 15 segments: the 4 PVs; the antra of
each PV; the anterior, posterior, and septal surfaces and roof and floor
of the LA; the mitral isthmus; and the LAA, as shown in Figure 1.
The presence of EAT, high DF sites, and CFAE sites in the PVs and
LA segments was assessed, and the correspondence between the EAT
locations and the high DF and CFAE sites was quantified.
Statistical Analysis
Continuous variables are expressed as the mean and 95% CI.
Distributions of the serum triglyceride, hs-CRP, and IL-6 levels were
skewed and are, therefore, expressed as median and interquartile
ranges; Mann-Whitney U test was used to analyze differences in
these variables between patients with PAF and those with PerAF.
Because the other continuous variables (including baseline values,
κ values, total agreement of EAT with high DF/CFAE, and DF/CFAE
values) were normally distributed, absolute differences (along with
95% CIs) were calculated and analyzed by Student t test. Fisher exact
probability test was used to compare the distribution of dichotomous
variables between PAF and PerAF patients. Simple linear regression
analysis was performed to test the correlation of total EAT and LA-EAT
with hs-CRP. Bland and Altman plots with 95% limits of agreement
were used to assess interobserver and intraobserver reproducibility of
EAT measurements. Agreement between the location of EAT and the
high DF and CFAE sites in the 15 LA and PV segments was assessed
by κ statistic, which was calculated by subtracting the proportion of
the readings expected to agree by chance, which we called Pe, from
the overall agreement, Po, and dividing the remainder by the number
Table 1. Demographics and Baseline Values in Overall Patient Series and Per Study Group
Variable
Total Patients (n=34)
PAF Group (n=16)
PerAF Group (n=18)
P Value
(PAF vs PerAF)
Age, y
56 (51–60)
55 (47–62)
56 (50–63)
0.73
Sex: male (%)
29 (85)
14 (88)
15 (83)
1.0
BMI, kg/m2
23.3 (22.2–24.4)
22.2 (20.6–23.9)
7 (44)
24.2 (22.7–25.8)
7 (39)
0.07
Hypertension (%)
14 (41)
HbA1c, %
5.4 (5.2–5.5)
5.3 (5.1–5.5)
5.5 (5.2–5.8)
1.0
0.28
HDL-C, mg/dL
59 (52–65)
64 (55–73)
54 (45–64)
0.12
LDL-C, mg/dL
113 (104–121)
119 (106–131)
107 (94–120)
0.18
Triglyceride, mg/dL
130 (99–157)
122 (66–151)
144 (113–165)
0.16
LAV, mL
41.4 (33.9–48.8)
38.1 (30.5–45.7)
46.5 (34.2–58.8)
0.24
LVEF, %
64.5 (61.7–67.4)
65.4 (62.7–68.1)
63.7 (58.6–68.8)
0.56
173.2 (151.5–194.8)
148.7 (119.9–177.6)
194.9 (164.2–225.5)
0.028
44.4 (37.4–51.3)
34.8 (26.6–43.0)
52.9 (44.2–61.5)
0.007
hs-CRP, ng/mL
500 (216–1280)
320 (120–660)
969 (307–1678)
0.008
IL-6, pg/mL
2.0 (1.0–2.9)
1.3 (0.8–2.4)
2.4 (1.7–3.2)
0.017
MMP-2, ng/mL
759 (712–807)
702 (645–759)
810 (739–881)
0.019
Echocardiographic measures
EAT
Total EAT, cm3
LA-EAT, cm3
Biomarkers of AF
PAF indicates paroxysmal atrial fibrillation; PerAF, persistent AF; BMI, body mass index; HbA1c, hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol;
LDL-C, low-density lipoprotein cholesterol; LAV, left atrial volume; LVEF, left ventricular ejection fraction; EAT, epicardial adipose tissue; LA-EAT, left atrium EAT; hs-CRP,
high-sensitivity C-reactive protein; IL-6, interleukin-6; MMP-2, matrix metalloproteinase-2.
Values shown are mean (95% CI), median (25th percentile, 75th percentile), or n (%).
 Nagashima et al Epicardial Adipose Tissue and DF 679
of cases in which agreement was not expected to occur by chance:
κ=(Po−Pe)/(1−Pe).
Kappa statistic values range from −1.0 to +1.0, with 0 indicating
chance agreement and +1.0 indicating perfect agreement. κ>0.75 implies excellent agreement, values 0.40 to 0.75 suggest fair-to-good
agreement, and values <0.4 imply poor agreement.27 P<0.05 was considered statistically significant. All statistical analyses except derivation of the κ statistics were performed with JMP 8 software (SAS
Institute, Cary, NC).
Results
Baseline Characteristics, Biomarker Levels, and
Echocardiographic Features of PAF and PerAF
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
Patients’ baseline characteristics, laboratory values, and
echocardiographic features of PAF and PerAF are shown in
Table 1. There were no differences in baseline characteristics,
standard blood chemistry values, or echocardiographic LA
volume and left ventricular ejection fraction between the 2
groups. EAT volumes were significantly greater in patients
with PerAF than in those with PAF (total EAT, 194.9 cm3 [95%
CI, 164.2–225.5] versus 148.7 cm3 [95% CI, 119.9–177.6];
absolute difference, 46.1 cm3 [95% CI, 5.4–86.9]; P=0.028;
LA-EAT, 52.9 cm3 [95% CI, 44.2–61.5] versus 34.8 cm3
[95% CI, 26.6–43.0]; absolute difference, 18.1 cm3 [95%
CI, 5.4–30.7]; P=0.007). Serum hs-CRP, IL-6, and MMP-2
levels were significantly higher in patients with PerAF than
in those with PAF (hs-CRP, 969 ng/mL [307–1678] versus
320 ng/mL [120–660]; P=0.008; IL-6, 2.4 pg/mL [1.7–3.2]
versus 1.3 pg/mL [0.8–2.4]; P=0.017; MMP-2, 810 ng/mL
[95% CI, 739–881] v­ ersus 702 ng/mL [95% CI, 645–759];
absolute difference, 108 ng/mL [95% CI, 19–197]; P=0.019).
In addition, the correlations between EAT volumes and hsCRP levels were as follows: r=0.57, P=0.0005 for total EAT;
r=0.55, P=0.0009 for LA-EAT.
Reproducibility of the EAT Measurements
EAT volume was measured by 2 independent operators.
Interobserver reproducibility was high according to Bland
and Altman analysis, with bias of 1.8 cm3 and 95% CI for the
estimate of the bias of −0.9 to 4.6 for total EAT and bias of
0.7 cm3 and 95% CI of −1.2 to 2.5 for LA-EAT. Intraobserver
reproducibility was also high, with bias of 0.1 cm3 and 95% CI
of −2.3 to 2.6 for total EAT and bias of 0.9 cm3 and 95% CI of
−2.0 to 0.1 for LA-EAT.
Relationships Between Location of EAT and
High DF and CFAE Sites
Before PVI, 343±98 data points per patient were acquired
for the creation of NavX-based LA and PV maps. Representative examples of a DF map and a CFAE map merged
with EAT are shown in Figure 2. The incidences of EAT,
Figure 2. A representative case of paroxysmal atrial fibrillation showing the
distribution of epicardial adipose tissue
(EAT), high-dominant frequency (DF)
sites, and complex fractionated atrial
electrogram (CFAE) sites on NavX maps.
On the DF map, the areas in bright purple are defined as the high DF sites with
a frequency of >8 Hz. Whereas on the
CFAE map, the white areas are defined
as continuous CFAEs with a mean fractionation interval (FI) <50 ms, and the
areas in red, orange, yellow, green, or
blue are variable CFAEs having an FI of
50 to 120 ms. Note that EAT is located at
the anterior portion of the right superior
pulmonary vein (PV) and left superior PV,
antra of 4 PVs, roof and floor of the left
atrium (LA), mitral isthmus, LA appendage; these locations are adjacent to the
high DF sites rather than the CFAE sites.
CFE indicates complex fractionated
electrogram.
680 Circ Arrhythm Electrophysiol August 2012
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
high DF, and CFAE at each LA/PV segment are shown in
Table 2. The incidences of EAT were >60% at the antra of
the left superior PV, left inferior PV, and right superior PV;
the anterior surface, roof, and floor of the LA; the mitral
isthmus; and the LAA. Similarly, high DF sites were often
located at the antra of each PV; the anterior surface, roof,
and floor of the LA; the mitral isthmus; and the LAA. In contrast, CFAEs were often located at the antra of each PV and
throughout the LA body (roof, floor, septal portion, anterior
and posterior surfaces of the LA, and the LAA). Most high
DF sites corresponded to the EAT sites; the CFAE sites did
not correspond as frequently with the EAT sites (Table 2).
Overall, the average total agreement of EAT with high DFs
in LA/PV segments was significantly higher than that of
EAT with CFAEs (90.0% [95% CI, 87.5–92.5] versus 68.4%
[95% CI, 63.0–73.8]; absolute difference, 21.6% [95% CI,
16.6–26.6]; P<0.0001). There was also excellent agreement
between EAT locations and high DF sites (κ=0.77 [95% CI,
0.71–0.82]), but there was poor agreement between EAT
locations and CFAE sites (κ=0.22 [95% CI, 0.13–0.31]).
When comparing the prevalence with the total agreement
of EAT and high DFs or CFAEs between PAF and PerAF
patients, no difference per segment was noted according to
Fisher exact probability test (Tables 3 and 4). In fact, the
good agreement between EAT locations and high DF sites
and the poor agreement between EAT locations and CFAE
sites were similar between PAF and PerAF patients (high
DF sites, κ=0.75 [95% CI, 0.68–0.82] versus 0.78 [95%
CI, 0.69–0.87], respectively; absolute difference, 0.03
[95% CI, −0.08 to 0.15]; P=0.58; CFAEs, κ=0.21 [95% CI,
0.06–0.36] versus 0.23 [95% CI, 0.10–0.36], respectively;
absolute difference, 0.02 [95% CI, −0.17 to 0.21]; P=0.82).
DFs and CFAEs With and Without EAT
Mean DF in each LA/PVs segment corresponding to an EAT
site was significantly greater than mean DF in segments
without EAT (8.6 Hz [95% CI, 8.4–8.7] versus 6.4 Hz [95%
CI, 6.2–6.5]; absolute difference, 2.2 Hz [95% CI, 2.0–2.5];
P<0.0001). Mean CFE-mean value at each atrial segment
corresponding to an EAT site was significantly shorter than
at segments without EAT (95 ms [95% CI, 90–100] versus
127 ms [95% CI, 119–136]; absolute difference, 32 ms [95%
CI, 23–42]; P<0.0001).
Discussion
Major Findings
We quantified EAT of the LA and PVs in patients with AF
and evaluated its specific distribution. PerAF patients, in
comparison with PAF patients, were shown to have increased
total EAT and LA-EAT volumes and elevated biomarkers of
inflammation and collagen turnover. High DF sites, rather
than CFAE sites, corresponded to the EAT sites, and this overlap between EAT and high DF sites was observed in both PAF
and PerAF patients.
Presence of EAT and Its Influence on AF
EAT was observed most often at the antra of the left superior PV,
left inferior PV, and right superior PV; the anterior surface, roof,
and floor of the LA; the mitral isthmus; and the LAA, and these
sites corresponded to the high DF sites rather than the CFAE
Table 2. Segmental Distribution of EAT, High DFs, and CFAEs and Total Agreement in Distribution Among Total Patients (n=34)
High DFs
Sensitivity,
%
Specificity,
%
CFAEs
Segment
EAT (%)
Presence
(%)
Total Agreement
With EAT (%)
Presence
(%)
Sensitivity,
%
Specificity,
%
Total Agreement
With EAT, (%)
LSPV
15 (44)
16 (47)
81
89
29 (85)
13 (38)
62
67
22 (65)
LSPV antrum
23 (68)
22 (65)
83
73
27 (79)
26 (76)
83
36
23 (68)
LIPV
20 (59)
20 (59)
100
100
34 (100)
12 (35)
83
50
21 (62)
LIPV antrum
27 (79)
24 (71)
89
100
31 (91)
21 (62)
67
57
22 (65)
RSPV
18 (53)
16 (47)
100
89
32 (94)
12 (35)
67
55
20 (59)
RSPV antrum
28 (82)
23 (68)
82
100
29 (85)
25 (74)
79
50
25 (74)
RIPV
10 (29)
10 (29)
100
100
34 (100)
9 (26)
78
88
29 (85)
RIPV antrum
26 (34)
24 (71)
92
100
32 (94)
26 (76)
81
38
24 (71)
LA roof
33 (97)
31 (91)
100
25
31 (91)
27 (79)
100
14
28 (82)
LA septum
17 (50)
20 (59)
85
100
31 (91)
27 (79)
48
43
16 (47)
LA anterior
23 (68)
23 (68)
100
100
34 (100)
27 (79)
74
57
24 (71)
LA posterior
16 (47)
19 (56)
74
87
27 (79)
31 (91)
48
67
17 (50)
LA floor
27 (79)
26 (76)
100
88
33 (97)
31 (91)
77
0
24 (71)
Mitral isthmu
33 (97)
28 (82)
100
17
29 (85)
25 (74)
96
0
24 (71)
LAA
34 (100)
25 (74)
74
NA
25 (74)
30 (88)
89
NA
30 (88)
EAT indicates epicardial adipose tissue; DF, dominant frequency; CFAEs, complex fractionated atrial electrograms; LSPV, left superior pulmonary vein; LIPV, left inferior
PV; RSPV, right superior PV; RIPV, right inferior PV; LA, left atrium; LAA, LA appendage.
Agreement with EAT, correspondence between EAT distribution and high DF and CFAE distributions in each LA/PV segment.
The numbers and percentages of the patients are shown. The presence, patient number, and percentage of patient numbers divided by 34 total patients with high
DFs or CFAEs in each segment are given.
 Nagashima et al Epicardial Adipose Tissue and DF 681
Table 3. Segmental Distribution of EAT, High DFs, and CFAEs and Total Agreement in Distribution Among PAF Patients (n=16)
High DFs
Segment
EAT (%)
Presence
(%)
Sensitivity,
%
Specificity,
%
CFAEs
Total Agreement
With EAT (%)
Presence
(%)
Sensitivity,
%
Specificity,
%
Total Agreement
With EAT (%)
LSPV
7 (44)
6 (38)
83
80
13 (81)
6 (38)
50
60
9 (56)
LSPV antrum
9 (56)
10 (63)
89
71
13 (81)
14 (88)
100
29
11 (69)
LIPV
8 (50)
8 (50)
100
100
16 (100)
4 (25)
50
50
8 (50)
12 (75)
12 (75)
100
100
16 (100)
10 (63)
67
50
10 (63)
6 (38)
4 (25)
100
83
14 (88)
7 (44)
43
67
9 (56)
12 (75)
8 (50)
67
100
12 (75)
13 (81)
83
25
11 (69)
6 (38)
6 (38)
100
100
16 (100)
5 (31)
100
91
15 (94)
LIPV antrum
RSPV
RSPV antrum anterior
RIPV
RIPV antrum anterior
12 (75)
12 (75)
100
100
LA roof
16 (100)
15 (94)
100
0
16 (100)
11 (69)
75
50
11 (69)
14 (88)
12 (75)
100
0
12 (75)
LA septum
9 (56)
11 (69)
82
100
14 (88)
13 (81)
54
33
8 (50)
LA anterior
12 (75)
12 (75)
100
100
16 (100)
14 (88)
79
50
12 (75)
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
5 (31)
6 (38)
67
90
13 (81)
15 (94)
33
100
6 (38)
LA floor
11 (69)
10 (63)
100
83
15 (94)
15 (94)
67
0
10 (63)
Mitral isthmus
15 (94)
12 (75)
100
25
13 (81)
13 (81)
92
0
12 (75)
LAA
16 (100)
11 (69)
69
NA
11 (69)
14 (88)
88
NA
14 (88)
LA posterior
The presence, patient number, and percentage of patient numbers divided by 16 PAF patients with high DFs or CFAEs in each segment are given.
EAT indicates epicardial adipose tissue; DF, dominant frequency; CFAEs, complex fractionated atrial electrograms; PAF, paroxysmal atrial fibrillation; LSPV, left
superior pulmonary vein; LIPV, left inferior PV; RSPV, right superior PV; RIPV, right inferior PV; LA, left atrium; LAA, LA appendage.
sites. Recently, CFAEs have been thought to represent localized
reentry, AF drivers, wavefront collision, and conduction through
channels of functional block and pivot points.7 In contrast, high
DF sites indicate local synchronous high-frequency activations
that are likely to be closely linked to localized reentry or the AF
driver(s).8 Detailed analysis of the NavX-based DF and CFAE
maps showed that most CFAE activity occurred near a high
DF site and that the cores of the widely distributed continuous
CFAEs sites correlated with the high DF sites.23 Therefore,
overlap between the locations of EAT and the high DF
sites implies that EAT is most likely to harbor high-frequency
sites, producing a favorable condition for maintenance of AF.
Two potential mechanisms explaining the correspondence between EAT and high DF sites can be considered.
Table 4. Segmental Distribution of EAT, High DFs, and CFAEs and Total Agreement in Distribution Among PerAF Patients (n=18)
High DFs
Segment
LSPV
LSPV antrum
EAT (%)
Presence
(%)
Sensitivity,
%
Specificity,
%
CFAEs
Total Agreement
With EAT (%)
Presence
(%)
Sensitivity,
%
Specificity,
%
Total Agreement
With EAT (%)
8 (44)
10 (56)
80
100
16 (89)
7 (39)
71
73
13 (72)
14 (78)
12 (67)
79
75
14 (78)
12 (67)
71
50
12 (67)
LIPV
12 (67)
12 (67)
100
100
18 (100)
8 (44)
100
50
13 (72)
LIPV antrum
15 (83)
12 (67)
80
100
15 (83)
11 (61)
67
67
12 (67)
RSPV
12 (67)
12 (67)
100
100
18 (100)
5 (28)
100
46
11 (61)
RSPV antrum
16 (89)
15 (83)
94
100
17 (94)
12 (67)
75
100
14 (78)
4 (22)
4 (22)
100
100
18 (100)
4 (22)
50
86
14 (78)
RIPV antrum
RIPV
14 (78)
12 (67)
86
100
16 (89)
15 (83)
86
25
13 (72)
LA roof
17 (94)
16 (89)
100
50
17 (94)
15 (83)
100
33
16 (89)
8 (44)
9 (50)
89
100
17 (94)
14 (78)
43
50
8 (44)
12 (67)
LA septum
LA anterior
11 (61)
11 (61)
100
100
18 (100)
13 (72)
69
60
LA posterior
11 (61)
13 (72)
77
80
14 (78)
16 (89)
63
50
11 (61)
LA floor
16 (89)
16 (89)
100
100
18 (100)
16 (89)
88
0
16 (89)
Mitral isthmus
18 (100)
16 (89)
100
0
16 (89)
12 (67)
100
0
12 (67)
LAA
18 (100)
14 (78)
78
NA
14 (78)
16 (89)
89
NA
16 (89)
The presence, patient number, and percentage of patient numbers divided by 18 PerAF patients with high DFs or CFAEs in each segment are given.
EAT indicates epicardial adipose tissue; DF, dominant frequency; CFAEs, complex fractionated atrial electrograms; PerAF, persistent atrial fibrillation; LSPV, left
superior pulmonary vein; LIPV, left inferior PV; RSPV, right superior PV; RIPV, right inferior PV; LA, left atrium; LAA, LA appendage.
682 Circ Arrhythm Electrophysiol August 2012
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
First, several activated proinflammatory cytokines are
secreted from the EAT.15,16 Its proximity to atria without
fascial boundaries and its blood supply from the coronary
arteries suggest that EAT may interact locally with atrial
myocardium through endocrine and paracrine secretion
of proinflammatory cytokines.11–16 In fact, levels of serum
inflammatory cytokines, such as hs-CRP and IL-6, were
higher in our PerAF patients having a greater EAT volume than in our PAF patients. These cytokines released
from EAT may change the electrophysiological characteristics of atrial and PV cardiomyocytes, which in turn may
be the substrate for development of high DF sites, leading
to the progression of AF.28 It was previously reported that
serum MMP-2 is implicated in atrial remodeling.20 The
MMP-2 level was higher in our PerAF patients than in our
PAF patients. This increased MMP-2 level in patients with
greater EAT volume supports contribution of EAT to the
progression of atrial remodeling.
The second possibility is the relationship between ganglionated plexi (GP) and AF. EAT contains abundant GP, which
are a critical element responsible for the initiation and maintenance of AF.29 The GPs are located at the LA roof, medial to
the left superior PV and often extending to the medial aspect
of the LAA (left superior GP), the anterior portion to the right
superior PV and often extending from the anterior region to
the right inferior PV (right anterior GP), the inferior portions
of the left and right inferior PVs (left inferior GP and right
inferior GP), and within the ligament of Marshall (Marshall
tract GP).29 We found total agreement between high DF and
EAT to be >80% at 4 PVs and their antra except left superior PV, the anterior surface, the septal portion, roof, and floor
of the LA, and the mitral isthmus. Notably, these sites corresponded to the locations of GP. GP activation includes both
parasympathetic and sympathetic stimulation of the atria/
PVs adjacent to the GP. Parasympathetic stimulation shortens the action potential duration, and sympathetic stimulation
increases calcium loading and calcium release from the sarcoplasmic reticulum. The combination of the short action potential duration and longer calcium release induces triggered
firing resulting from delayed afterdepolarization of the atria/
PVs adjacent to the GP as manifested by the high DF sites.
Overlap between EAT and high DF locations was seen in both
the PAF and PerAF groups, indicating that cytokine secretion
or GP firing from EAT may be related to maintenance of AF,
regardless of the AF type. This agrees with the fact that extensive PVI and CFAE/DF-based substrate modification is clinically effective not only for PAF, but also for PerAF; extensive
circumferential PV isolation and ablation target sites for substrate modification include most of the EAT or GP sites.29
Study Limitations
Several study limitations must be considered. First, the study
involved a relatively small number of patients. We, therefore,
meticulously analyzed the distribution of EAT, high DF sites,
and CFAE sites on NavX maps by acquiring a large number
of data points. Second, the automatic CFE-mean algorithm is
a specific method for detecting CFAEs available only on the
NavX system. However, the algorithm has been confirmed
by catheter ablation results in many reports.22 Third, we do
not have sufficient data to show that the high DFs are generated directly by the EAT, proinflammatory cytokines secreted
locally from the EAT, or migration of the axon to the atrium
from the epicardial GPs. It is difficult to confirm a causal
relationship between EAT and high DFs on the basis of clinical observations. Finally, we did not assess the relationships
between EAT and mapped DF and CFAE sites after PVI. The
circumferential lesions of our extensive PVI-based catheter
ablation cover several high DF sites adjacent to EAT, and this
hinders us in clarifying the influence of EAT on the electrophysiological properties of the LA body in AF.
Conclusions
Increased EAT volume and elevated biomarkers of inflammation and collagen turnover are noted in PerAF patients compared with the measurements in PAF patients. High DF sites
are located adjacent to EAT sites. Thus, EAT may be involved
in the maintenance of AF.
Disclosures
None.
References
1. Calkins H, Brugada J, Packer DL, Cappato R, Chen SA, Crijns HJ, Damiano RJ Jr, Davies DW, Haines DE, Haissaguerre M, Iesaka Y, Jackman W, Jais P, Kottkamp H, Kuck KH, Lindsay BD, Marchlinski FE,
McCarthy PM, Mont JL, Morady F, Nademanee K, Natale A, Pappone
C, Prystowsky E, Raviele A, Ruskin JN, Shemin RJ; Heart Rhythm Society; European Heart Rhythm Association; European Cardiac Arrhythmia
Society; American College of Cardiology; American Heart Association;
Society of Thoracic Surgeons. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. A report of the
Heart Rhythm Society (HRS) Task Force on Catheter and Surgical Ablation of Atrial Fibrillation developed in partnership with the European
Heart Rhythm Association (EHRA) and the European Cardiac Arrhythmia
Society (ECAS); in collaboration with the American College of Cardiology (ACC), American Heart Association (AHA), and the Society of Thoracic Surgeons (STS). Endorsed and approved by the governing bodies of
the American College of Cardiology, the American Heart Association, the
European Cardiac Arrhythmia Society, the European Heart Rhythm Association, the Society of Thoracic Surgeons, and the Heart Rhythm Society.
Europace. 2007;9:335–379.
2. Haïssaguerre M, Sanders P, Hocini M, Takahashi Y, Rotter M, Sacher F,
Rostock T, Hsu LF, Bordachar P, Reuter S, Roudaut R, Clémenty J, Jaïs P.
Catheter ablation of long-lasting persistent atrial fibrillation: critical structures for termination. J Cardiovasc Electrophysiol. 2005;16:1125–1137.
3.O’Neill MD, Jaïs P, Takahashi Y, Jönsson A, Sacher F, Hocini M,
Sanders P, Rostock T, Rotter M, Pernat A, Clémenty J, Haïssaguerre M.
The stepwise ablation approach for chronic atrial fibrillation–evidence for
a cumulative effect. J Interv Card Electrophysiol. 2006;16:153–167.
4. Schmitt C, Estner H, Hecher B, Luik A, Kolb C, Karch M, Ndrepepa
G, Zrenner B, Hessling G, Deisenhofer I. Radiofrequency ablation of
complex fractionated atrial electrograms (CFAE): preferential sites of
acute termination and regularization in paroxysmal and persistent atrial
fibrillation. J Cardiovasc Electrophysiol. 2007;18:1039–1046.
5. Atienza F, Calvo D, Almendral J, Zlochiver S, Grzeda KR, MartínezAlzamora N, González-Torrecilla E, Arenal A, Fernández-Avilés F,
Berenfeld O. Mechanisms of fractionated electrograms formation in the
posterior left atrium during paroxysmal atrial fibrillation in humans. J Am
Coll Cardiol. 2011;57:1081–1092.
6. Moe GK, Mendez C. Basis of pharmacotherapy of cardiac arrhythmias.
Mod Concepts Cardiovasc Dis. 1962;31:739–744.
7. Gerstenfeld EP, Lavi N, Bazan V, Gojraty S, Kim SJ, Michele J. Mechanism of complex fractionated electrograms recorded during atrial fibrillation in a canine model. Pacing Clin Electrophysiol. 2011;34:844–857.
8. Takahashi Y, Sanders P, Jaïs P, Hocini M, Dubois R, Rotter M, Rostock
T, Nalliah CJ, Sacher F, Clémenty J, Haïssaguerre M. Organization of
 Nagashima et al Epicardial Adipose Tissue and DF 683
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
frequency spectra of atrial fibrillation: relevance to radiofrequency catheter ablation. J Cardiovasc Electrophysiol. 2006;17:382–388.
9. Wanahita N, Messerli FH, Bangalore S, Gami AS, Somers VK, Steinberg
JS. Atrial fibrillation and obesity–results of a meta-analysis. Am Heart J.
2008;155:310–315.
10. Watanabe H, Tanabe N, Watanabe T, Darbar D, Roden DM, Sasaki S,
Aizawa Y. Metabolic syndrome and risk of development of atrial fibrillation: the Niigata preventive medicine study. Circulation. 2008;117:
1255–1260.
11. Nagashima K, Okumura Y, Watanabe I, Nakai T, Ohkubo K, Kofune T,
Kofune M, Mano H, Sonoda K, Hirayama A. Association between epicardial adipose tissue volumes on 3-dimensional reconstructed CT images and recurrence of atrial fibrillation after catheter ablation. Circ J.
2011;75:2559–2565.
12. Shin SY, Yong HS, Lim HE, Na JO, Choi CU, Choi JI, Kim SH, Kim JW,
Kim EJ, Park SW, Rha SW, Park CG, Seo HS, Oh DJ, Kim YH. Total and
interatrial epicardial adipose tissues are independently associated with left
atrial remodeling in patients with atrial fibrillation. J Cardiovasc Electrophysiol. 2011;22:647–655.
13.Thanassoulis G, Massaro JM, O’Donnell CJ, Hoffmann U, Levy D,
Ellinor PT, Wang TJ, Schnabel RB, Vasan RS, Fox CS, Benjamin EJ. Pericardial fat is associated with prevalent atrial fibrillation: the Framingham
Heart Study. Circ Arrhythm Electrophysiol. 2010;3:345–350.
14.Al Chekakie MO, Welles CC, Metoyer R, Ibrahim A, Shapira AR,
Cytron J, Santucci P, Wilber DJ, Akar JG. Pericardial fat is independently associated with human atrial fibrillation. J Am Coll Cardiol.
2010;56:784–788.
15. Iacobellis G, Barbaro G. The double role of epicardial adipose tissue as
pro- and anti-inflammatory organ. Horm Metab Res. 2008;40:442–445.
16.Mazurek T, Zhang L, Zalewski A, Mannion JD, Diehl JT, Arafat H,
Sarov-Blat L, O’Brien S, Keiper EA, Johnson AG, Martin J, Goldstein BJ,
Shi Y. Human epicardial adipose tissue is a source of inflammatory mediators. Circulation. 2003;108:2460–2466.
17.Chung MK, Martin DO, Sprecher D, Wazni O, Kanderian A, Carnes
CA, Bauer JA, Tchou PJ, Niebauer MJ, Natale A, Van Wagoner DR.
C-reactive protein elevation in patients with atrial arrhythmias: inflammatory mechanisms and persistence of atrial fibrillation. Circulation.
2001;104:2886–2891.
18. Engelmann MD, Svendsen JH. Inflammation in the genesis and perpetuation of atrial fibrillation. Eur Heart J. 2005;26:2083–2092.
19. Issac TT, Dokainish H, Lakkis NM. Role of inflammation in initiation
and perpetuation of atrial fibrillation: a systematic review of the published
data. J Am Coll Cardiol. 2007;50:2021–2028.
20.Okumura Y, Watanabe I, Nakai T, Ohkubo K, Kofune T, Kofune M,
Nagashima K, Mano H, Sonoda K, Kasamaki Y, Hirayama A. Impact
of biomarkers of inflammation and extracellular matrix turnover on the
outcome of atrial fibrillation ablation: importance of matrix metalloproteinase-2 as a predictor of atrial fibrillation recurrence. J Cardiovasc Electrophysiol. 2011;22:987–993.
21. Ujino K, Barnes ME, Cha SS, Langins AP, Bailey KR, Seward JB, Tsang
TS. Two-dimensional echocardiographic methods for assessment of left
atrial volume. Am J Cardiol. 2006;98:1185–1188.
22.Lin YJ, Tai CT, Kao T, Chang SL, Lo LW, Tuan TC, Udyavar AR,
Wongcharoen W, Hu YF, Tso HW, Tsai WC, Chang CJ, Ueng KC, Higa S,
Chen SA. Spatiotemporal organization of the left atrial substrate after circumferential pulmonary vein isolation of atrial fibrillation. Circ Arrhythm
Electrophysiol. 2009;2:233–241.
23.Lin YJ, Tsao HM, Chang SL, Lo LW, Hu YF, Chang CJ, Tsai WC,
Suenari K, Huang SY, Chang HY, Wu TJ, Chen SA. Role of high dominant frequency sites in nonparoxysmal atrial fibrillation patients: insights
from high-density frequency and fractionation mapping. Heart Rhythm.
2010;7:1255–1262.
24. Park JH, Park SW, Kim JY, Kim SK, Jeoung B, Lee MH, Hwang C, Kim
YH, Kim SS, Pak HN. Characteristics of complex fractionated atrial electrogram in the electroanatomically remodeled left atrium of patients with
atrial fibrillation. Circ J. 2010;74:1557–1563.
25. Verma A, Lakkireddy D, Wulffhart Z, Pillarisetti J, Farina D, Beardsall M,
Whaley B, Giewercer D, Tsang B, Khaykin Y. Relationship between complex fractionated electrograms (CFE) and dominant frequency (DF) sites
and prospective assessment of adding DF-guided ablation to pulmonary
vein isolation in persistent atrial fibrillation (AF). J Cardiovasc Electrophysiol. 2011;22:1309–1316.
26. Koruth JS, Heist EK, Danik S, Barrett CD, Kabra R, Blendea D, Ruskin J,
Mansour M. Accuracy of left atrial anatomical maps acquired with a multielectrode catheter during catheter ablation for atrial fibrillation. J Interv
Card Electrophysiol. 2011;32:45–51.
27. Fleiss JL. The measurement of interrater agreement. In: Statistical Methods for Rates and Proportions. New York, NY: John Wiley; 1981:212–236.
28. Lin YK, Chen YJ, Chen SA. Potential atrial arrhythmogenicity of adipocytes: implications for the genesis of atrial fibrillation. Med Hypotheses.
2010;74:1026–1029.
29. Po SS, Nakagawa H, Jackman WM. Localization of left atrial ganglionated plexi in patients with atrial fibrillation. J Cardiovasc Electrophysiol.
2009;20:1186–1189.
CLINICAL PERSPECTIVE
Numerous studies have shown that increased epicardial adipose tissue (EAT) volume is associated with an increased prevalence of atrial fibrillation (AF) and that EAT volume can predict the development of AF. Indeed, EAT is a source of several
inflammatory mediators, which might be linked to the development of AF. However, the mechanisms for how EAT impacts
the atria remain unclear. High dominant frequency (DF) and complex fractionated atrial electrogram sites are well known to
be key electrophysiological parameters reflecting microreentrant circuits or sites of focal-firing that drive AF. These parameters are often used by electrophysiologists as electrogram-based target sites for AF ablation. Hypothesizing that EAT might
promote high DF and complex fractionated atrial electrogram sites, we sought to investigate whether EAT locations correspond to high DF or complex fractionated atrial electrogram sites during AF through a unique 3-dimentional merge process,
ie, the NavX-based complex fractionated atrial electrogram and DF left atrial map with 3-dimensional reconstructed EAT.
The present study found a good association between high DF sites during AF and the distribution of EAT. Furthermore, the
serum level of the high-sensitivity C-reactive protein, which is a kind of inflammatory cytokine, was elevated in patients with
remodeled atria and increased EAT volume. Those findings indicated that EAT was involved in the development of high DF
sites because of the secretion of inflammatory cytokines and the induction of firing of ganglionated plexi, which are abundant
in EAT. The present article should provide new insight into the key drivers of AF.
Does Location of Epicardial Adipose Tissue Correspond to Endocardial High Dominant
Frequency or Complex Fractionated Atrial Electrogram Sites During Atrial Fibrillation?
Koichi Nagashima, Yasuo Okumura, Ichiro Watanabe, Toshiko Nakai, Kimie Ohkubo,
Masayoshi Kofune, Hiroaki Mano, Kazumasa Sonoda, Takafumi Hiro, Mizuki Nikaido and
Atsushi Hirayama
Downloaded from http://circep.ahajournals.org/ by guest on June 18, 2017
Circ Arrhythm Electrophysiol. 2012;5:676-683; originally published online July 6, 2012;
doi: 10.1161/CIRCEP.112.971200
Circulation: Arrhythmia and Electrophysiology is published by the American Heart Association, 7272 Greenville
Avenue, Dallas, TX 75231
Copyright © 2012 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-3149. Online ISSN: 1941-3084
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circep.ahajournals.org/content/5/4/676
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Arrhythmia and Electrophysiology can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for which
permission is being requested is located, click Request Permissions in the middle column of the Web page
under Services. Further information about this process is available in the Permissions and Rights Question and
Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Arrhythmia and Electrophysiology is online at:
http://circep.ahajournals.org//subscriptions/