DOI: 10.1161/CIRCEP.111.967612
A Novel Assessment of the Temporal Variation in the Fractionated
Electrograms Using a Histogram Analysis of the Local Fractionation Interval
in Persistent Atrial Fibrillation Patients
Running title: Lin et al.; Histogram analysis of atrial signals
Yenn-Jiang Lin, MD1,2*; Kazuyoshi Suenari, MD1,3*; Men-Tzung Lo, PhD4; Chen Lin, PhD4;
Wan-Hsin Hsieh, PhD4; Shih-Lin Chang, MD1,2; Li-Wei Lo, MD1,2; Yu-Feng Hu, MD1,2;
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Chen-Chuan Cheng, MD5; Yasuki Kihara, MD3; Tze-Fan Chao, MD
D1,2, ; Be
Beny
ny H
Hartono,
arto
ar
tono
to
no, M
no
MD1,2;
Tsu-Juey Wu, MD2,6; Wei-Shiang Lin, MD7; Ke-Hsin Hsu, MS4; Ambr
Ambrose
brrose S. K
Kibos,
ibos, M
ib
MD1;
12
N d E Huang,
Norden
H
PhD4; Shih-Ann
Shih A Chen,
Ch MD1,2
1
Division off Cardiology, Taipei Veterans General Hospital; 2School of Medicine, Institute of Clinic
Clinical
c Medicine,
Cardiovascular
Taipei,Taiwan;
s
scular
Research Center, National Yang-Ming University, Taipei,
Taiwan; 3Dept of Cardiovascular
Card
d
Medicine, Hiroshima
University Graduate School of Biomedical Sciences, Hiroshima, Japan; 4Res
H
Research
s
Center
for Adaptive
National
Central
v Data Analysis & Center for Dynamical Biomarkers and Translational Medicine, Nati
ve
i
University,, Chungli; 5Division of Cardiology, Chi-Mei Medical Center, Tainan; 6Division of Cardiology,
Cardii
Dept
of Medicine,
e Taichung Veterans General Hospital, Taichung, 7Division of Cardiology, Dept of Inter
e,
Internal
r Medicine,
National Defense Medical Center & Tri-Service General Hospital, Taipei, Taiwan
*contributed equally
Address for Correspondence:
Shih-Ann Chen, MD
Division of Cardiology
Taipei Veterans General Hospital
201, Sec. 2, Shih-Pai Road
Taipei, Taiwan
Tel: +886-2-2875-7156
Fax: +886-2-2873-5656
E-mail: [email protected]
Journal Subject Codes: [22] Ablation/ICD/surgery
1
DOI: 10.1161/CIRCEP.111.967612
Abstract:
Background - The characteristics of atrial electrograms associated with atrial fibrillation (AF)
termination are controversial. We investigated the electrogram characteristics that indicate
procedural AF termination during continuous complex fractionated electrogram (CFE) ablation.
Methods and Results - Fifty-two consecutive persistent AF patients (47 males, 54 ± 9 years)
who underwent electrogram-based catheter ablation in the left atrium and coronary sinus
following pulmonary vein isolation (PVI) were enrolled. The intracardiac bipolar atrial
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
ysis (>6s
6s),
6s
) (2)
electrogram recordings were characterized by (1) fractionation intervall (FI) analy
analysis
(>6s),
sttogram)
m).
m)
) S
it
kurtosis (shape of the FI histogram), and (3) skewness (asymmetry of the F
FII hi
histogram).
Sites
F (mean FI ”60 ms) were targeted, and AF was terminated in 20 patients (38%
FEs
%
showing CFEs
(38%)
V The conventional CFE sites (mean ” 120 ms) in the patients with AF termina
VI.
a
after the PVI.
termination
h
2
exhibited higher
median kurtosis (2.69 (IQR 2.03–3.46) vs. 2.35 (IQR 1.79–2.48), P = 0.02
0.024)
h
and higher CFE-mean interval (102.7±19.8 vs. 87.7±15.0, P = 0.008) than the patients with
without
ti
M
th AF termination
t
i ti sites
it hhad
d hi
h median
di kurtosis
k t i th
AF termination.
Moreover,
the
higher
than th
the ttargeted
sites without AF termination (5.13 (IQR 3.51–6.47) vs. 4.18 (IQR 2.91–5.34), P < 0.01) in
patients with procedural termination. In addition, patients with AF termination had a higher sinus
rhythm maintenance rate after a single procedure than patients without AF termination (Log-rank
test, P = 0.007).
Conclusions - A kurtosis analysis using the FI histogram may be a useful tool for identifying the
critical substrate for persistent AF and potential responders to catheter ablation.
Key words: atrial fibrillation; catheter ablation; electrophysiology mapping; histogram; left
atrium
2
DOI: 10.1161/CIRCEP.111.967612
Introduction
Atrial fibrillation (AF) is the most common sustained tachyarrhythmia encountered in clinical
practice.1 Pulmonary vein isolation (PVI) is the mainstream catheter ablation technique for atrial
fibrillation (AF).2,3 For longer durations of AF, substrate modification with complex fractionated
electrogram (CFE) ablation is necessary for patients who do not respond to PVI. The
development of automated analysis algorithms for electrogram fractionation is important for a
reproducible and objective assessment of this technique.4-6 However, most algorithms are based
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
domain electrogra
ams
m such
on the mean fractionation interval (FI) between the deflection of time-domain
electrograms
ARTO
AR
TO sys
y tem.
m T
as the CFE-mean of the NavX system or shortest complex interval of the C
CARTO
system.
The
n the FIs acquired by these modalities may be important for the algorithms of CF
F
variation in
CFEs.
Therefore, if the local FIs are not normally distributed, the mean FI has limited clinical
t due to the temporal variation. We hypothesized that in addition to the conventional
ty
conventt
applicability
m the temporal variation in the annotated FIs may provide important informatio
m,
o for
FI algorithm,
information
t
iti l CFE
t d was tto evaluate
l t a novell
determining th
the ffeatures
off critical
CFEs. Th
The purpose off thi
this study
signal analysis concept based on the histogram of the FI values acquired by the CFE-mean
algorithm including the kurtosis (i.e., shape of the FI distribution) and skewness (i.e., asymmetry
of the FI distribution).7
Methods
Patient Characteristics
Fifty-two consecutive persistent AF patients who underwent radiofrequency ablation guided by a
3D anatomical mapping system (NavX, St. Jude. Medical, St. Paul, MN, USA) were studied
retrospectively. Persistent AF (including long-standing persistent AF) was defined according to
the consensus statement on catheter and surgical ablation of AF.8 At the beginning of the
3
DOI: 10.1161/CIRCEP.111.967612
procedure, all patients presented with incessant AF. Patients with paroxysmal AF or persistent AF
terminated by PVI were excluded.
Electrophysiological Study
After informed consent was obtained, each patient underwent an electrophysiological study and
catheter ablation in a fasting state. All antiarrhythmic drugs except amiodarone were
discontinued for at least 5 half-lives before the procedure. The 3D electroanatomic mapping
method was performed as described previously.9 Mapping was performed with an irrigated
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
iinserted
3.5-mm tip deflectable catheter (EPT, Boston Scientific Corporation, Natick, MA, USA)) in
or an ad
ddi
d tiion
onal
into the left atrium (LA ) alongside the transseptal sheath without the need ffor
additional
i The 3D geometry of the LA was then generated using the NavX system.
ite.
puncture site.
blation
Catheter A
Ablation
e ablation of AF involved the following steps:
er
The catheter
1 PV isolation: After a successful transseptal procedure, continuous circumferen
n
Step 1,
circumferential
i li th
i ht and
d lleft
ft PV ostia
ti guided
id d bby th
X system
t were created
t d using
i a
lesions encircling
the right
the N
NavX
3.5-mm irrigated-tip ablation catheter (EPT, Boston Scientific Corporation, Natick, MA, USA)
as described previously.9 Radiofrequency energy was applied continuously while repositioning
the catheter tip every 40 s with a target temperature of 35–40°C and maximum power of 25–30
W in the power control mode. After the circumferential lesion set was completed, the ipsilateral
superior and inferior PVs were mapped carefully by recording with a circular catheter (Spiral,
AF Division, St. Jude Medical). Supplementary ablation applications were applied along the
circumferential lines close to the earliest ipsilateral PV potentials.
Step 2, Continuous CFE site ablation: If AF did not stop after PVI, an additional
CFE-guided substrate ablation was performed on the basis of the post-PVI CFE maps. Since a
4
DOI: 10.1161/CIRCEP.111.967612
prospective randomized study demonstrated that the empirical right atrial (RA) CFE ablation
does not improve long-term AF-free survival,10 the CFE ablation was confined to the continuous
CFEs (mean FI ” 60 ms) in the LA and proximal coronary sinus (CS).11,12 The right atrial CFEs
were not targeted. The end points of the CFE site ablation were to obtain prolonged cycle length,
eliminate the CFEs (mean FI >120 ms), or abolish the local fractionated potentials (bipolar
voltage <0.05 mV). The end point of step 2 was the elimination of all continuous CFEs in the LA
and CS. If AF terminated during the linear ablation through the CFE sites, a complete linear
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
ablation to the mitral annulus or nearest ablation line was performed to
o prevent pr
ppro-arrhythmias.
o-arrh
hyt
yth
R) by el
lectr
t ic
i card
rdio
rd
iov
io
If AF did not stop after step 2, it was restored to sinus rhythm (SR)
electric
cardioversion.
i
irectional
conduction block and dissociated PV activity were confirmed during S
PV-LA bidirectional
SR.
Additional ablation at the sites off the residual PV potentials was aapplied
pplied from the atrial sidee of
r
rum
the PV antrum
using the electrogram-guided approach to obtain entrance block.
3 Non-PV ectopies: After SR was restored from AF either by procedural AF
Step 3,
termination or electric
l t i cardioversion,
di
i mapping
i andd ablation
bl ti were only
l applied
li d to
t spontaneously
t
initiating focal atrial tachycardias and non-PV ectopies that initiated AF.13 First, we placed
multiple catheters in the LA (ablation and spiral catheters), superior vena cava (SVC)-RA
(duodecapolar catheter), and CS to identify the location of the AF-initiating ectopy. Second, we
used a mapping catheter to identify the earliest activation site of the ectopy. If any non-PV
ectopies initiating AF from the SVC were identified, the SVC was isolated guided by the circular
catheter recordings from the SVC–RA junction.
In this study, if the AF became organized during steps 1–2, electroanatomic mapping and
ablation were performed to terminate the organized tachycardia. The AF termination was defined
as AF restored to SR during the ablation in steps 1–2 without cardioversion. All patients
5
DOI: 10.1161/CIRCEP.111.967612
underwent an RA cavotricuspid isthmus ablation with an 8-mm-tip ablation catheter. No AF
inducibility tests using any ȕ-agonists or pacing maneuvers were performed after the ablation
procedure.
Signal Recording and Analysis
After acquiring the LA geometry, a 3.5-mm-tip catheter was selected as the “roving” catheter for
sequential contact mapping as described previously,9,14,15. Regarding the protocol for signal
sampling in the LA and CS, we divided the LA into 10 parts as follows: the high and low
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
al isthmus, anterior
or aand
n
anteroseptal wall, high and low posterior wall, medial and lateral mitral
sites
ites weree
it
posterior roof, and right and left pulmonary vein–atrial junction.14 At least 10 sit
determined
d in each region. The numbers of sites in each region were similar and nearly equ
equally
distributed.. To avoid high-density mapping in the procedural termination sites, the distancee of
any paired adjacent mapping sites <5 mm was detected by the off-line software based on th
the
h 3D
location.9 Therefore,
T
we selected the nearest site to analyze the characteristics of the procedural
procedd
termination sites.
it
Regarding fractionation mapping, the NavX mapping parameters were set to the CFE-mean,
which is an interval analysis algorithm that measures the average index of the fractionation at
each site and produces a color map representative of the CFE distribution. The settings used have
been described previously:5 a refractory period of 30 ms, peak-to-peak sensitivity from 0.05–0.1
mV, and duration of 10 ms. A CFE was defined as mean FI ” 120 ms, while non-CFEs were
defined as FI>120 ms. Sites of continuous CFEs were the most fractionated sites with a local
mean FI of ”60 ms and recording duration of >6 s. The color annotation was set to a range of
colors with an FI of ”60 ms shown in white to an FI of >120 ms shown in purple. The white
areas in the NavX map represent the regions with continuous CFEs. Sites with shorter mean FIs
6
DOI: 10.1161/CIRCEP.111.967612
indicated a high degree of temporal stability of the fractionated electrograms. The consistency of
the fractionated electrograms over time was previously validated.5
The assessed electrograms in each patient were acquired and characterized by the FI
analysis. Furthermore, the statistical properties of the FI, including the kurtosis and skewness of
FI distribution over a 6-s window, were assessed. Briefly, kurtosis measures the shape of the
distribution of the fractionated intervals within the window beyond simply using means and
standard deviations (Figure 1A). Kurtosis indicates the relationship between each of the FIs to
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
the mean; the higher the kurtosis, the less probability that FIs deviate from the mean. Mea
Meanwhile,
ean
ea
t s if tth
he F
Is ca
skewness assesses the degree of asymmetry of the distribution, which indicate
indicates
the
FIs
can be
moo
sufficientlyy represented by the mean or if they should be represented by the median and mode
(Figure 1B).
). The higher the skewness, the larger the differences between the mean FIs and their
n modes. In addition, positive or negative skewness means that the mode and m
nd
medians and
median
a lower or higher than the mean of the FIs, respectively. These features of bipolar
bipol
of the FIs are
electrogram recordings
di
iin patients
ti t with
ith and
d without
ith t procedural
d l AF ttermination
i ti were analyzed
l
and compared. In the patients with AF termination, we selected the procedural AF termination
site for the signal analysis. However, if no AF termination site was located on the prior CFE
mapping sites, we adopted the most adjacent mapping sites <5 mm away for the data analysis.
Before exporting the recording data, the quality of the contact electrograms and CFE maps was
confirmed by 2 independent physicians experienced in signal analysis.
Follow-up of AF Recurrence
After discharge, the patients underwent follow-up—2 weeks after the catheter ablation and every
1–3 months thereafter—at our cardiology clinic or with the referring physicians; antiarrhythmic
drugs were prescribed for 8 weeks to prevent any early recurrence of AF. During each follow-up
7
DOI: 10.1161/CIRCEP.111.967612
after the ablation, 24-h Holter monitoring and/or cardiac event recording were performed for 1
week to determine the cause of the clinical symptoms. Recurrence of atrial arrhythmia, including
atrial tachycardia, atrial flutter, or AF, was defined as an episode lasting more than 1 min and
confirmed by ECG 2 months after the ablation. Patients with AF recurrence were asked to
undergo a repeat ablation procedure after a blanking period of 2 months.16 After the blanking
period, we stopped the antiarrhythmic drugs in all patients without any AF recurrence.
Statistical Analysis
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Data were presented as the mean value ± standard deviation (SD) if normally
ormally distributed.
d. T
The
ge: m
ed
dian, Q
Q1
1–
non-normally distributed data were presented as quartiles (interquartilee rang
range:
median,
Q1–Q3).
a
ariables
continn
Nominal variables
were compared with Fisher’s exact test. The normally distributed continuous
w compared using the Student’s t-test and one-way ANOVA, whereas the
variables were
a distributed variables were compared using the Mann-Whitney U test and
ally
non-normally
W
Wallis
f
regression analysis was used to
Kruskal–Wallis
test with a Bonferroni
correction. Logistic regression
th predictors
di t off AF termination.
t
i ti The
Th predicted
di t d probability
b bilit off each
h subject
bj t was
determine the
calculated with mean FI and mean FI plus median kurtosis.” The area under the ROC curve was
used to quantify the discrimination of each model. DeLong’s method was used to compare the
areas under the ROC curves of these 2 models. The normal distribution of the FI values was
evaluated using the Shapiro–Wilk test. Kaplan–Meier curves were performed to determine the
SR maintenance rates after the last procedure; the rates were then compared using the log-rank
test. The level of statistical significance was set at P < 0.05 (2-tailed). If a Bonferroni correction
was applied, a corrected P value of 0.017 was used.
Results
Patient Characteristics and Catheter Ablation
8
DOI: 10.1161/CIRCEP.111.967612
The clinical characteristics of the study population are shown in Table 1. Patients without AF
termination had higher incidences of long-standing persistent AF and hypertension than those
with AF termination (P < 0.01). Patients without AF termination had a larger LA diameter and
lower left ventricular ejection fraction on 2D echocardiography than patients with AF
termination.
In the initial ablation step, PVI was performed successfully and electrical isolation was
confirmed in all patients. However, PVI did not terminate the AF in all patients. In the next step,
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
R was
wa
w
CFE ablation terminated the AF in 20 patients (38%); in the other 32 patients (62%), SR
on, 6 ppatients
attientts (3
(30%
% of the
restored by electrical cardioversion. In the patients with AF termination,
th AF termination) were restored to SR by CFE ablation around the CS. The residual
resii
patients with
a and/or non-pulmonary vein triggers were subsequently eliminated in all patie
als
e
PV potentials
patients.
All patientss successfully underwent a cavotricuspid isthmus ablation.
a
aracteristics
Signal Characteristics
at the Patient Level
th signal
i l characteristics
h
t i ti off all
ll CFEs
CFE off the
th patients
ti t with
ith and
d without
ith t AF
Table 2 shows the
termination at the patient level. There were no significant differences in the median FI or skewness
between the patients with and without AF termination (P > 0.05). The patients with AF termination
had higher median kurtosis [2.69 (2.03–3.46) vs. 2.35 (1.79–2.48), P=0.024] and longer mean FI
(102.7 ± 19.8 vs. 87.7 ± 15.0, P=0.008) than those in the patients without AF termination.
Signal Characteristics at Individual CFE Sites
Table 3 summarizes the signal characteristics of the targeted continuous CFE sites (FI ”60 ms) in
the patients with and without AF termination. According to the Shapiro–Wilk test, both patients
with and without AF procedural termination had low incidences of normal distributions of FI
histograms not only in the conventional CFE sites (” 120 ms) (10 ± 8% vs. 14 ± 14% per patient,
9
DOI: 10.1161/CIRCEP.111.967612
respectively, P = 0.45), but also in the continuous CFE sites (” 60 ms, 6 ± 8% vs. 8 ± 9% per
patient, respectively, P = 0.64). The patients with AF termination had higher kurtosis for the
continuous CFE sites than patients without AF termination. In the patients with AF termination,
the AF termination sites had significantly higher kurtosis than the non-termination sites (P <
0.01). However, the skewness was not significantly different between the targeted sites with and
without AF termination (P > 0.05).
Prediction of Procedural Termination
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
edu
d r AF
Figures 2 and 3 show the signal analysis results in selected cases with and without proce
procedural
i the
tthhe patients
pattie
i ntts w
termination. The histogram analysis revealed that the AF termination sites in
with AF
terminationn exhibited maximal CFEs, which corresponds to high kurtosis; moreover, the
distributionn of the maximal CFEs was compatible with the high kurtosis area. Furthermore,,
e with high kurtosis in the non-continuous CFE areas were observed in both
es
several sites
terminationn and non-termination patients, indicating the presence of stable and organized
ti it with
ith a llong cycle
l llength.
th
fibrillatory activity
Logistic regression analysis determined the predictors of AF termination; the odds ratios for
CFF-mean and median kurtosis were 12.07 and 30.84, respectively (P = 0.013 and 0.004,
respectively). The predicted probability of each subject was calculated using CFE-mean and
CFE-mean plus median kurtosis. The area under the ROC curve was used to quantify the
discrimination of each model. The area under the curve was greater for CFE-mean plus kurtosis
(AUC = 0.835) than for CFE-mean only (AUC = 0.68) with a sensitivity of 75% and specificity
of 89% (Figure 4).
Long-term Follow-up
The rate of SR maintenance was 50% (N = 26) with a mean follow-up duration of 14 ± 8 months
10
DOI: 10.1161/CIRCEP.111.967612
(median, 12 months) after a single procedure. Meanwhile, among the patients with recurrence of
atrial arrhythmias (N = 26), 17 (65%), 4 (15%), 3 (12%), and 2 (8%) had AF, typical atrial flutter,
atypical atrial flutter, atrial tachycardia, respectively. The SR maintenance rate was significantly
lower in the patients without AF termination than that in the patients with AF termination
(Log-rank test, P = 0.007, Figure 5). Furthermore, after a mean of 1.4 ± 0.6 procedures per
patient (median, 2 procedures), the rate of SR maintenance was 67% (mean, 17 ± 8 months,
median, 16 months, N = 35), the SR maintenance rate after the last procedure in the patients
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
ts with AF termination
terminaati
tio
without AF termination was significantly lower than that in the patients
(Log-rank test, P = 0.038).
n
Discussion
d
dings
Main Findings
histogrr
To the bestt of our knowledge, this is the first report addressing the novel concept of histogram
p
analysis forr substrate mapping in patients with persistent
AF. The results of this studyy
demonstrate that histogram analysis from the annotated FI data provides information on the
temporal variations of the annotated FI in the fractionation analysis. The lower temporal
variability of the continuous CFEs (FI ” 60 ms) evidenced by the high kurtosis associated with
AF termination indicates the reliability of the mean FI. Furthermore, the response to the
continuous CFE ablation was higher. However, in sites with low kurtosis, conventional methods
using the mean fractionation interval may be inappropriate for identifying the critical sites with a
low procedural AF termination rate. This study demonstrates the adjunctive role of histogram
analysis in the conventional mean interval-based algorithm for prediction termination.
Predicting AF Termination during Catheter Ablation
Since PVI alone is ineffective for treating patients with persistent AF,8 it is necessary to perform
11
DOI: 10.1161/CIRCEP.111.967612
additional substrate modification following complete PVI. Since several investigators report a
correlation between procedural AF termination and better clinical outcomes in patients with
AF,17-19 most laboratories define procedural AF termination as an endpoint in persistent AF
patients. However, achieving procedural AF termination, especially in patients with persistent AF,
may require more extensive procedures including CFE modification and/or linear ablation.20-23
Although the results of the present study also revealed high SR maintenance rates in patients
with AF termination, AF termination was achieved in only 38% of this population even though
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
we targeted the continuous CFE sites following the complete isolation of all 4 PVs. A rec
recent
ecen
ec
e
iindicator
dicatorr of
di
o
study by Natale et al. found that procedural AF termination may not bee a goodd in
20
o
However, in that study, the incidence of AF termination directly to S
SR
long-term outcomes.
was low (2%).
%). Despite converting to AT in 56% of their study population, most patients stil
still
l
required electrical
e
ectrical
cardioversion following further extensive ablation. These results indicatee that
the current algorithm based on the CFE-mean is not optimal. Thus, it could be difficult to
differentiate
the unimportant
t the
th responsible
ibl CFEs
CFE from
f
th
i
t t ones.
Implications of the Novel Histogram Analysis Method
A recent clinical study (CFAE-AF) demonstrates that ablating certain grades of CFE increases
AF cycle length irrespective of the order of CFE ablation; this indicates that different grades of
CFEs play different roles.24 The different roles of CFEs were also revealed by MAP recordings,
demonstrating the characteristics of the responsible sites, which differ from non-local and
far-field signals.25 At present, most automatic algorithms are based on the mean interval during
AF. The present study is the first to use histogram analysis to evaluate the shape of the FI
distribution within a 6-s window. For individual mapping sites, higher kurtosis indicates a lower
probability that FIs deviate from the mean. Moreover, since skewness assesses the degree of
12
DOI: 10.1161/CIRCEP.111.967612
asymmetry of a distribution, it indicates whether the FIs can be sufficiently represented by the
mean or median or mode. The temporal variability of the FIs can be used as an adjunctive tool
for identifying the critical atrial substrate. At the patient level, the characteristics of continuous
CFEs in patients with AF termination were different from those without AF termination. In
patients with AF termination, high kurtosis in the maximal CFEs corresponded to the reliability
of the mean FI values of these regions (Figures 2C and 2D). On the contrary, low kurtosis in the
continuous CFEs indicated larger variations in the FIs, less accuracy of the mean FI values for
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
ariations in the FIs
FIIs and
a
an
detecting the CFEs, and a lower response to catheter ablation. These variations
p li
pp
licati
tion off th
ti
the cycle
non-normally distributed datasets in the histogram analysis may limit the app
application
e automatic algorithms of CFEs for defining target sites. Low kurtosis of the CF
ed
F
length-based
CFE
a the complexity of the fibrillation signals as well as the difficulty of isolatingg the
ated
sites indicated
d demonstrate that tthe
signals for precise analysis. Furthermore, the results of the present study
o of kurtosis with the conventional CFE-mean can improve the predictive accur
on
r
combination
accuracy
of
t
i ti comparedd with
ith the
th use off the
th CFE-mean
CFE
l alone.
l
Th in
i patients
procedural AF termination
value
Thus,
with persistent AF undergoing catheter ablation, histogram analysis may provide additional
information regarding the reliability of the automatic algorithm of the FIs and predict the
response to catheter ablation.
Study Limitations
The present study is limited because it was a single-center retrospective study with a relatively
small number of patients. In addition, the priority of CFE ablation was not based on the kurtosis.
A further prospective study is warranted to confirm the role of these regions. The morphology of
the discrete electrograms during AF may provide some useful information for AF ablation.25,26
However, we only investigated the variation in the intervals and not the electrogram morphology,
13
DOI: 10.1161/CIRCEP.111.967612
which requires further study. We did not analyze the FI histograms before PVI. Therefore, our
findings could be affected by the elimination of a critical area for AF maintenance along the PVI
line as reported previously.9 Finally, the atrial de-bulking effect may exist when targeting the
next CFE site; this is the universal limitation faced when investigating the characteristics of CFE
sites.
Conclusions
In persistent AF patients with AF termination, continuous CFEs (mean FI”
” 60 ms)) are
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
term
rmin
rm
inat
in
atio
at
ion.
io
n. A kkurtosis
u
characterized by a high stationary status compared to patients without AF te
termination.
analysis using
ing the FI histogram may be a useful tool for identifying the critical substrate for
A and potential responders to catheter ablation. Continuous CFEs with higher
persistent AF
t
tability
between intervals are shown to be an important characteristic of ablation sites
temporal stability
p
that cause procedural
AF termination in persistent AF patients. In patients without AF
n, low kurtosis in the continuous CFEs indicated larger
g variations in the FIs,, less
termination,
accuracy of the mean FI values for detecting the CFEs.
Funding Sources: This work was supported was supported by National Science Council (NSC;
Taiwan, ROC), Grant No 100-2221-E-008-008-MY2 and No 99-2628-B-010-024-MY3, joint
foundation of Cathay General Hospital and National Central University (NCU), Grant No
CNJRF-99CGH-NCU-A3, joint foundation of Taipei Veterans General Hospital and NCU,
VGHUST100-G1-4-3, and NSC supports for the Center for Dynamical Biomarkers and
Translational Medicine, National Central University, Taiwan ( NSC 100-2911-I-008-001)
Conflict of Interest Disclosures: None.
References:
1. Nattel S. New ideas about atrial fibrillation 50 years on. Nature. 2002;415:219-226.
2. Chen SA, Hsieh MH, Tai CT, Tsai CF, Prakash VS, Yu WC, Hsu TL, Ding YA, Chang MS.
14
DOI: 10.1161/CIRCEP.111.967612
Initiation of atrial fibrillation by ectopic beats originating from the pulmonary veins:
electrophysiological characteristics, pharmacological responses, and effects of radiofrequency
ablation. Circulation. 1999;100:1879-1886.
3. Haissaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, Quiniou G, Garrigue S, Le
Mouroux A, Le Metayer P, Clementy J. Spontaneous initiation of atrial fibrillation by ectopic
beats originating in the pulmonary veins. N Engl J Med. 1998;339:659-666.
4. Verma A, Novak P, Macle L, Whaley B, Beardsall M, Wulffhart Z, Khaykin Y. A prospective,
multicenter evaluation of ablating complex fractionated electrograms (CFEs) during atrial
fibrillation (AF) identified by an automated mapping algorithm: acute effects on AF and efficacy
as an adjuvant strategy. Heart Rhythm. 2008;5:198-205.
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
5. Lin YJ, Tai CT, Kao T, Chang SL, Wongcharoen W, Lo LW, Tuan TC, Udyavar AR, Chen YJ,
al electrog
grams dur
urin
ur
i
Higa S, Ueng KC, Chen SA. Consistency of complex fractionated atrial
electrograms
during
atrial fibrillation. Heart Rhythm. 2008;5:406-412.
Marin
ne JE,
JE Berger
Berg
Be
rger
rg
er RD
6. Scherr D, Dalal D, Cheema A, Cheng A, Henrikson CA, Spragg D, Marine
RD,
a
Calkins H, Dong J. Automated detection and characterization of complex fractionated atria
atrial
m in human left atrium during atrial fibrillation. Heart Rhythm. 2007;4:1013-10
ms
electrograms
2007;4:1013-1020.
R Jenkins JM, DiCarlo LA. Intraventricular electrogram analysis for ventricula
a
7. Throne RD,
ventricular
tachycardiaa detection: statistical validation. Pacing Clin Electrophysiol. 1990;13:1596-160
1990;13:1596-1601.
H Brugada J, Packer DL, Cappato R, Chen SA, Crijns HJ, Damiano RJ Jr., Davies
Davv
8. Calkins H,
es DE, Haissaguerre M, Iesaka Y, Jackman W, Jais P, Kottkamp H, Kuck KH, Lin
DW, Haines
Lindsay
hli ki FE
C th PM
d F
d
K
t l A
BD, Marchlinski
FE, M
McCarthy
PM, M
Montt JL
JL, M
Morady
F, N
Nademanee
K, N
Natale
A, P
Pappone C,
Prystowsky E, Raviele A, Ruskin JN, Shemin RJ. 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. Heart Rhythm. 2007;4:816-861.
9. 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.
10. Oral H, Chugh A, Good E, Crawford T, Sarrazin JF, Kuhne M, Chalfoun N, Wells D,
Boonyapisit W, Gadeela N, Sankaran S, Kfahagi A, Jongnarangsin K, Pelosi F, Jr., Bogun F,
Morady F. Randomized evaluation of right atrial ablation after left atrial ablation of complex
fractionated atrial electrograms for long-lasting persistent atrial fibrillation. Circ Arrhythm
Electrophysiol. 2008;1:6-13.
11. Ndrepepa G, Schneider MA, Karch MR, Weber S, Schreieck J, Schomig A, Schmitt C.
Pulmonary vein internal electrical activity does not contribute to the maintenance of atrial
fibrillation. Pacing Clin Electrophysiol. 2003;26:1356-1362.
15
DOI: 10.1161/CIRCEP.111.967612
12. Ndrepepa G, Schneider MA, Karch MR, Weber S, Schreieck J, Zrenner B, Schmitt C. Impact
of atrial fibrillation on the voltage of bipolar signals acquired from the left and right atria. Pacing
Clin Electrophysiol. 2003;26:862-869.
13. Lin WS, Tai CT, Hsieh MH, Tsai CF, Lin YK, Tsao HM, Huang JL, Yu WC, Yang SP, Ding
YA, Chang MS, Chen SA. Catheter ablation of paroxysmal atrial fibrillation initiated by
non-pulmonary vein ectopy. Circulation. 2003;107:3176-3183.
14. Chang SL, Tai CT, Lin YJ, Wongcharoen W, Lo LW, Tuan TC, Udyavar AR, Chang SH, Tsao
HM, Hsieh MH, Hu YF, Chen YJ, Chen SA. Biatrial substrate properties in patients with atrial
fibrillation. J Cardiovasc Electrophysiol. 2007;18:1134-1139.
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
15. Chang SL, Tsao HM, Lin YJ, Lo LW, Hu YF, Tuan TC, Suenari K, Tai CT, Li CH, Chao TF,
Recur
urre
ur
r n of
re
Lin YK, Tsai CF, Wu TJ, Chen SA. Characteristics and Significance off Veryy Earlyy Recurrence
ol. 20
ol
2011
11;2
11
;22:
;2
2:11
2:
1193
11
93-1
93
-1
119
Atrial Fibrillation After Catheter Ablation. J Cardiovasc Electrophysiol.
2011;22:1193-1198.
ari K
Huan
angg SY
an
S
16. Lo LW, Tsao HM, Lin YJ, Chang SL, Hu YF, Tsai WC, Tuan TC, Suena
Suenari
K, Hu
Huang
SY, Tung
S Tai CT, Ueng KC, Li CH, Chao TF, Wu TJ, Chen SA. Different patterns of atrial
atrr
NH, Higa S,
l
remodelingg after catheter ablation of chronic atrial fibrillation. J Cardiovasc Electrophysiol
Electrophysiol.
8
85-393.
2011;22:385-393.
C Verma A, Di Biase L, Ching CK, Patel D, Barrett C, Martin D, Rong B, Fahm
m TS,
17. Elayi CS,
Fahmy
Khaykin Y, Hongo R, Hao S, Pelargonio G, Dello Russo A, Casella M, Santarelli P, Potenzaa D,
M
R, Arruda M, Schweikert RA, Natale A. Ablation for longstanding perm
m
Fanelli R, Massaro
permanent
l
lation:
results from a randomized study comparing three different strategies. Hea
a
atrial fibrillation:
Heart
008 5 1658 1664
Rhythm. 2008;5:1658-1664.
18. Lin YJ, Tai CT, Chang SL, Lo LW, Tuan TC, Wongcharoen W, Udyavar AR, Hu YF, Chang
CJ, Tsai WC, Kao T, Higa S, Chen SA. Efficacy of additional ablation of complex fractionated
atrial electrograms for catheter ablation of nonparoxysmal atrial fibrillation. J Cardiovasc
Electrophysiol. 2009;20:607-615.
19. O'Neill MD, Wright M, Knecht S, Jais P, Hocini M, Takahashi Y, Jonsson A, Sacher F,
Matsuo S, Lim KT, Arantes L, Derval N, Lellouche N, Nault I, Bordachar P, Clementy J,
Haissaguerre M. Long-term follow-up of persistent atrial fibrillation ablation using termination
as a procedural endpoint. Eur Heart J. 2009;30:1105-1112.
20. Elayi CS, Di Biase L, Barrett C, Ching CK, al Aly M, Lucciola M, Bai R, Horton R, Fahmy
TS, Verma A, Khaykin Y, Shah J, Morales G, Hongo R, Hao S, Beheiry S, Arruda M, Schweikert
RA, Cummings J, Burkhardt JD, Wang P, Al-Ahmad A, Cauchemez B, Gaita F, Natale A. Atrial
fibrillation termination as a procedural endpoint during ablation in long-standing persistent atrial
fibrillation. Heart Rhythm. 2010;7:1216-1223.
21. Komatsu Y, Uno K, Otomo K, Nagata Y, Taniguchi H, Ogura K, Egami Y, Takayama K,
Kakita K, Iesaka Y. Atrial defibrillation threshold as a novel predictor of clinical outcome of
16
DOI: 10.1161/CIRCEP.111.967612
catheter ablation for persistent atrial fibrillation. Europace. 2011;13:213-220.
22. Lo LW, Tai CT, Lin YJ, Chang SL, Udyavar AR, Hu YF, Ueng KC, Tsai WC, Tuan TC,
Chang CJ, Kao T, Tsao HM, Wongcharoen W, Higa S, Chen SA. Predicting factors for atrial
fibrillation acute termination during catheter ablation procedures: implications for catheter
ablation strategy and long-term outcome. Heart Rhythm. 2009;6:311-318.
23. Yoshida K, Chugh A, Good E, Crawford T, Myles J, Veerareddy S, Billakanty S, Wong WS,
Ebinger M, Pelosi F, Jongnarangsin K, Bogun F, Morady F, Oral H. A critical decrease in
dominant frequency and clinical outcome after catheter ablation of persistent atrial fibrillation.
Heart Rhythm. 2010;7:295-302.
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
24. Hunter RJ, Diab I, Tayebjee M, Richmond L, Sporton S, Earley MJ and Schilling RJ.
Characterization of fractionated atrial electrograms critical for maintenance of atrial fibrillation:
l).
) Circ Arrhythm
a randomized, controlled trial of ablation strategies (the CFAE AF trial).
Electrophysiol. 2011;4:622-629..
iyazaki
ki S
Sach
cher
ch
er F
25. Narayan SM, Wright M, Derval N Jadidi A, Forclaz A, Nault I, Miyazaki
S, Sa
Sacher
F,
Bordachar P, Clementy J, Jais P, Haissaguerre M and Hocini M. Classifying fractionated
m in human atrial fibrillation using monophasic action potentials and activation
ms
electrograms
e
H
mapping: evidence
for localized drivers, rate acceleration, and nonlocal signal etiologies. Heart
0
011;8:244-53.
Rhythm. 2011;8:244-53.
o EJ, Biviano AB, Whang W, Vest JA, Gambhir A, Einstein AJ, Garan H. Differe
e
26. Ciaccio
Differences
persistt
in repeatingg patterns of complex fractionated left atrial electrograms in longstanding persistent
l
lation
as compared with paroxysmal atrial fibrillation. Circ Arrhythm Electrophysiol.
Electrophyy
atrial fibrillation
0 477
2011;4:470-477.
Table 1. Baseline Characteristics
Variables
Age, year-old
Male, N (%)
Long-standing AF, N (%)
Associated condition
Hypertension, N (%)
Diabetes mellitus, N (%)
Hyperlipidemia, N (%)
Coronary artery disease, N (%)
Echocardiographic findings
Left atrial diameter (mm)
Left ventricular ejection fraction (%)
AF termination (-)
(N=32)
54±7
30 (94%)
26 (81%)
AF termination (+)
(N=20)
55±11
17 (85%)
3 (15%)
23 (72%)
5 (16%)
11 (34%)
8 (25%)
4 (20%)
1 (5%)
4 (20%)
2 (10%)
<0.01
0.24
0.21
0.17
46±6
52±12
41±5
59±7
<0.01
<0.01
17
P value
0.70
0.28
<0.01
DOI: 10.1161/CIRCEP.111.967612
Table 2. Differences in the Signal Characteristics between Patients with and without AF
Termination (Sites with CFE<120 msec)
Variables
Mean FI
Median FI (25%-75%, msec)
Kurtosis (median, 25%-75%)
Skewness (median, 25%-75%)
FI= fractionation interval
Patient without
AF termination
(N=32)
87.7±15.0
72.7 (59.2-81.7)
2.35 (1.79-2.48)
1.50 (1.35-1.57)
Patient with
AF termination
(N=20)
102.7±19.8
82.5 (84.9-118.6)
2.69 (2.03-3.46)
1.52 (1.45-1.73)
P value
0.008
0.053
0.024
0.306
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Table 3. Summary of the Signal Characteristics in the CFE-targeted Ablation
blati
tion
on S
Sit
Sites
ites
it
es ((”
(”60
”60
60 msec)
ms
Variables
Sites in patients
without AF
termination
(N=559)
Patients with AF termination
Targeted sites
without immediate
termination
(N=178)
Targeted sites with
immediate
termination
(N=20)
P value
FI (median, 25%-75%,
2
msec)
55.4 (50.3-59.8)
49.5 (46.5-59.1)
53.2 (44.2-58.1)
0.27
Kurtosis (median,
dian, 25%-75%)
3.72 (2.95-4.55)
4.18 (2.91-5.34) *
5.13 (3.51-6.47) *†
<0.01
median 25%-75%)
Skewness (median,
11.24
24 (0
94-1 54)
(0.94-1.54)
11.23
23 (0
81-1 58)
(0.81-1.58)
11.65
65 (1
09-2 06)
(1.09-2.06)
0.09
* P<0.01 when compared to sites without termination in non AF-termination patients.
† P<0.01 when compared to sites without immediate termination in termination patients
Figure Legends:
Figure 1: Histogram Analysis. (A) The formula for the kurtosis of the univariate data Y1,
Y2…YN from the FI data. Compared with a Gaussian distribution (i.e., normal distribution), the
probability distribution with large kurtosis (see green line) has a higher distinct amplitude, fatter
tails, and declines rapidly; meanwhile, that with small kurtosis (see red line) has a more rounded
peak and thinner tails. (B) The formula for skewness for the univariate data Y1, Y2…YN from the
18
DOI: 10.1161/CIRCEP.111.967612
FI data. A skewness of zero indicates a perfect symmetric distribution, while a positive value
means a broader tail to the right compared to that on the left. ȸ is the mean, s is the standard
deviation, and N is the number of data points. FI: fractionation interval.
Figure 2: Sample Signal Analysis in a Case with AF Termination. (A), (B) The CFEs and
kurtosis map, respectively. The kurtosis map was constructed from the values of the FI histogram
analysis at each site. The most fractionated areas are at the low posterior wall and mitral isthmus.
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
ed the AF. (C
((C)) Thee
CFE-targeted ablation at the low posterior wall (red asterisk) terminated
togram att tthis
hiis ssite
it
histogram analysis of the FI over 6 s at the AF termination site (a). Thee hist
histogram
h
kurtosis with a sharp peaked distribution and positive skewness. This site
exhibited high
m
exhibited a continuous fractionated signal of 50 ms. (D) The point-by-point relationships am
among
t
tosis,
the FI, kurtosis,
and skewness off all CFE sites. High kurtosis wa
wass observed at the maximal CFE
s complex fractionated electrograms.
s:
sites. CFEs:
Figure 3: Sample Signal Analysis in a Case without AF Termination. (A), (B) The CFEs and
kurtosis map, respectively. The most fractionated areas are at the high anteroseptal wall near the
right PVI line with low kurtosis areas (black asterisk). CFE-targeted ablation at that site did not
terminate the AF. There was also a high kurtosis area at the anterior wall near the left atrial
appendage in the non-continuous CFE area. (C) The histogram analysis of the FI over 6 s at the
continuous CFE site (a). The histogram at this site exhibits low kurtosis and positive skewness;
meanwhile, this site exhibits a continuous fractionated signal of 53 ms. (D) The point-by-point
relationships among the FIs, kurtosis, and skewness of all CFE sites. There kurtosis was not high
at the maximal CFE sites.
19
DOI: 10.1161/CIRCEP.111.967612
Figure 4: ROC curve analysis for the prediction of procedural AF termination at the patient level
by using “CFE-mean” and “CFE-mean plus median kurtosis.” The probability score was used to
determine the optimal model for predicting AF termination.
Figure 5: Single and Multiple Procedure Success Rates. (A) Kaplan–Meier event-free survival
curve after a single catheter ablation attempt. (B) Kaplan–Meier event-free survival curve after
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
the last ablation catheter attempt.
20
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
A Novel Assessment of the Temporal Variation in the Fractionated Electrograms Using a
Histogram Analysis of the Local Fractionation Interval in Persistent Atrial Fibrillation Patients
Yenn-Jiang Lin, Kazuyoshi Suenari, Men-Tzung Lo, Chen Lin, Wan-Hsin Hsieh, Shih-Lin Chang,
Li-Wei Lo, Yu-Feng Hu, Chen-Chuan Cheng, Yasuki Kihara, Tze-Fan Chao, Beny Hartono, Tsu-Juey
Wu, Wei-Shiang Lin, Ke-Hsin Hsu, Ambrose S. Kibos, Norden E. Huang and Shih-Ann Chen
Downloaded from http://circep.ahajournals.org/ by guest on June 17, 2017
Circ Arrhythm Electrophysiol. published online July 25, 2012;
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/early/2012/06/23/CIRCEP.111.967612
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 Answerdocument.
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/