JACC: CLINICAL ELECTROPHYSIOLOGY
VOL. 1, NO. 4, 2015
ª 2015 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION
ISSN 2405-500X/$36.00
PUBLISHED BY ELSEVIER INC.
http://dx.doi.org/10.1016/j.jacep.2015.04.015
Mechanistic Comparison of
“Nearly Missed” Versus
“On-Target” Rotor Ablation
Masatoshi Yamazaki, MD, PHD,*y Uma Mahesh R. Avula, MD,* Omer Berenfeld, PHD,* Jérôme Kalifa, MD, PHD*
JACC: CLINICAL ELECTROPHYSIOLOGY CME
This article has been selected as the month’s JACC: Clinical Electrophy-
CME Objective for This Article: Upon completion of this activity, the
siology CME activity, available online at www.jacc-electrophysiology.org
learner should be able to discuss: 1) the role of rotors in the maintenance
by selecting the CME on the top navigation bar.
of atrial fibrillation; 2) why electrogram frequency is a useful atrial
electrogram analytical features to locate rotor regions; and 3) how point
Accreditation and Designation Statement
ablation may differentially modulate AF maintenance depending on its
exact location with respect to high frequency rotors.
The American College of Cardiology Foundation (ACCF) is accredited by
the Accreditation Council for Continuing Medical Education (ACCME) to
provide continuing medical education for physicians.
CME Editor Disclosure: CME Editor Smit Vasaiwala, MD, has nothing
to declare.
The ACCF designates this Journal-based CME activity for a maximum
of 1 AMA PRA Category 1 Credit(s). Physicians should only claim credit
Author Disclosures: This work was supported by the American Heart
commensurate with the extent of their participation in the activity.
Association Grant-in-Aid 13GRNT16820063 to Dr. Kalifa; NHLBI R01HL118304 to Dr. Berenfeld; and the Suzuken Memorial Foundation and
Method of Participation and Receipt of CME Certificate
the Grand-in-Aid Research Activity Start-up 25893090, scientific research
(C) 15K09077 of the Ministry of Education, Culture, Sports, Science and
To obtain credit for JACC: Clinical Electrophysiology CME, you must:
Technology, Japan, to Dr. Yamazaki. Dr. Berenfeld is a Co-founder,
1. Be an ACC member or JACC subscriber.
Scientific Officer, and shareholder of Rhythm Solutions, Inc.; has received
2. Carefully read the CME-designated article available online and in this
research support and a grant from Medtronic; and has received research
issue of the journal.
3. Answer the post-test questions. At least 2 out of the 3 questions
provided must be answered correctly to obtain CME credit.
support from St. Jude Medical. All other authors have reported that they
have no relationships relevant to the contents of this paper to disclose.
Drs. Yamazaki and Avula contributed equally to this work.
4. Complete a brief evaluation.
5. Claim your CME credit and receive your certificate electronically by
Medium of Participation: Print (article only); online (article and quiz).
following the instructions given at the conclusion of the activity.
CME Term of Approval
Issue Date: August 2015
Expiration Date: July 31, 2016
From the *Center for Arrhythmia Research, Cardiovascular Research Center, Department of Internal Medicine, Division of Cardiology, University of Michigan, Ann Arbor, Michigan; and the yResearch Institute of Environmental Medicine, Nagoya University,
Nagoya, Japan. This work was supported by the American Heart Association Grant-in-Aid 13GRNT16820063 to Dr. Kalifa; NHLBI
R01-HL118304 to Dr. Berenfeld; and the Suzuken Memorial Foundation and the Grand-in-Aid Research Activity Start-up
25893090, scientific research (C) 15K09077 of the Ministry of Education, Culture, Sports, Science and Technology, Japan, to Dr.
Yamazaki. Dr. Berenfeld is a Co-founder, Scientific Officer, and shareholder of Rhythm Solutions, Inc.; has received research
support and a grant from Medtronic; and has received research support from St. Jude Medical. All other authors have reported
that they have no relationships relevant to the contents of this paper to disclose. Drs. Yamazaki and Avula contributed equally to
this work.
Manuscript received March 13, 2015; revised manuscript received April 14, 2015, accepted April 16, 2015.
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
Mechanistic Comparison of “Nearly Missed”
Versus “On-Target” Rotor Ablation
ABSTRACT
OBJECTIVES This study used advanced optical mapping techniques to examine atrial fibrillation (AF) dynamics before
and after 2 distinct electrogram-based ablation strategies: complex fractionated atrial electrograms (CFAEs) and
DFmax/rotor ablation.
BACKGROUND Among the electrogram analytical features proposed to unravel the atrial regions that perpetuate AF,
CFAEs, highest dominant frequency sites (DFmax), and, more recently, phase analysis-enabled rotor mapping have
received the largest attention. Still, the mechanisms by which these approaches modulate AF dynamics and lead to AF
termination are unknown.
METHODS In Langendorff-perfused sheep hearts, AF was maintained by the continuous perfusion of acetylcholine and
high-resolution endocardial-epicardial optical videos were recorded from the left atrial free wall and the posterior left
atrium. Then, DFmax/rotor regions (n ¼ 7), or CFAE regions harboring the highest wavebreak density (HWD) (n ¼ 5), were
targeted with a 4F ablation catheter (5 to 15 W, 30 to 60 s/point). Thereafter, we examined the changes in AF dynamics
and whether AF terminated.
RESULTS DFmax/rotor point ablation resulted in a significant decrease in DFmax values. In 2 animals AF terminated,
whereas in the remaining 5 animals the post-ablation DFmax domain remained in the vicinity of its pre-ablation location.
However, after HWD/CFAEs density ablation, DFmax values did not change, AF did not terminate, and post-ablation
DFmax domains relocated from the left atrial free wall to the pulmonary vein–posterior left atrium region. In another
group of hearts (n ¼ 12), we observed that upon a progressive increase in acetylcholine concentration—mimicking the
acute electrophysiological changes occurring after ablation—3-dimensional rotors drifted from one atrial region to
another along large gradients of myocardial thickness.
CONCLUSIONS “On-target” DFmax/rotor ablation leads to the annihilation of the fibrillation-driving rotor. This
translates into large decreases in AF frequency or AF termination. In contrast, “nearly missed” HWD/CFAEs ablation
spares the fibrillation-driving rotor, and set the stage for rotor drift along large myocardial thickness gradients.
(J Am Coll Cardiol EP 2015;1:256–69) © 2015 by the American College of Cardiology Foundation.
E
lectrogram-based ablation consists of a de-
(3,4,8). Previously, our group has extensively described
tailed
electro-
that AF-perpetuating rotors activate the atrial muscle
grams during atrial fibrillation (AF) so as to
at an exceedingly high frequency of excitation (10–14).
identify fibrillation-maintaining regions. According
We showed that waves emanating from rapid reentrant
to the so-called stepwise ablation approach (1,2),
sources generate intense wavebreak activity in regions
electrogram-based ablation is performed after the isola-
located at the border between the DFmax domain and
tion of the pulmonary vein and posterior left atrium (PV-
surrounding domains (15); and that this relatively small
PLA). Among the many electrogram analytical features
band of high wavebreak density (HWD) corresponds to
that have been suggested to detect driver regions, com-
the regions where CFAEs are recorded. We showed
plex fractionated atrial electrograms (CFAEs), highest
experimentally that the DFmax domain encompasses
dominant frequency (DFmax ablation), and, more
the rotor region, whereas the DFmax and CFAEs do-
recently, phase analysis-enabled rotor detection have re-
mains are distinct, albeit adjacent (15). In humans, the
ceived the largest attention (3–8). In particular, rotor
recent successes of rotor ablation (6,7,9) suggest that
mapping studies in humans have suggested that the
the DFmax/rotor regions are key targets, but it is still un-
visualization and targeting of sustained reentrant AF
clear whether CFAE regions are also optimal sites of
sources may improve ablation success (6,7,9). However,
ablation. More generally, the electrophysiological mech-
whether rotor ablation approaches overlap with other
anisms by which any of these ablation approaches may
strategies such as DFmax or CFAEs ablation is unknown
succeed in terminating AF remain to be investigated.
analysis
of
intra-cardiac
257
258
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
ABBREVIATIONS
We present an experimental study in isolated
(Figure 1A) (16). The LAFW endocardial view spanned
AND ACRONYMS
hearts maintained in AF and mapped with a
an area representing approximately 30% of the free
high
ACh = acetylcholine
AF = atrial fibrillation
CFAE = complex fractionated
atrial electrogram
DFmax = highest dominant
frequency
HWD = high wavebreak density
LAFW = left atrial free wall
PVJ = pulmonary vein and
posterior left atrium junction
PV-PLA = pulmonary vein and
posterior left atrium
endocardial-
wall (Figure 1B). At each optical mapping field-of-
epicardial optical mapping apparatus. We pre-
view selected to record a video, high resolution
sent a detailed examination of AF dynamics
photographic snapshots of the endocardium and
before
electrogram/
epicardium were also obtained (Figure 1). They
mapping-based ablation strategies: DFmax/
served to superimpose endocardial and epicardial
rotor (“on-target”) and HWD/CFAEs (“nearly
patterns of electrical impulse propagation to the
missed”). We show that “on-target” DFmax/
corresponding left atrial anatomy anatomical back-
rotor ablation and “nearly missed” HWD/CFAEs
ground as previously (17). Then, a bolus of 15 ml
ablation lead to largely distinct outcomes and
Di-4-ANEPPS (10 mg/ml) was injected into the
resolution
and
simultaneous
after
2
distinct
perfusate to record voltage-sensitive fluorescence
AF dynamics.
changes (500 to 1,000 frames/s, 5-s videos, 80 80
pixels). Also, 10 m mol/ml of blebbistatin was used to
METHODS
reduce motion artifacts.
RI = regularity index
SW = scroll wave
WB = wavebreak
LANGENDORFF-PERFUSED
SHEEP
HEART
AND STRETCH-INDUCED AF MODEL. All an-
VIDEO ANALYSIS: DF AND REGULARITY INDEX ANALYSIS,
imal experiments were performed according
SINGLE-PIXEL ELECTROCARDIOGRAPHY. Videos of the
to the University of Michigan Committee on Use and
PLA and LAFW together with bipolar electrograms of
Care of Animals and the National Institutes of Health
the LAFW, left atrium-pulmonary vein junction (PVJ),
guidelines. Twenty-four sheep (35 to 40 kg) were
right atrial appendage (RAA), and coronary sinus
used as follows: on-target DFmax/rotor point ablation
made possible a precise characterization of DF dis-
protocol, n ¼ 7; nearly missed HWD/CFAEs point
tribution. The construction of average DF maps—
ablation protocol, n ¼ 5; and acetylcholine (ACh)
which are an average of 5 consecutive maps obtained
protocol, n ¼ 12. Sheep were anesthetized with an
at a 1-minute interval—allowed for a precise delinea-
intravenous bolus injection of propofol (5 to 10 mg/kg).
tion of the DFmax domain. Then, the rotor core’s
Then, hearts were excised and Langendorff-perfused
(X,Y) coordinates, or alternatively the DFmax domain
with warm oxygenated Tyrode’s solution (pH 7.4;
center region’s (X,Y) coordinates were noted as
95% O 2, 5% CO 2, and 36 C to 38 C). As previously
reference for the positioning of the ablation catheter
(14), we perforated the intra-atrial septum and
tip during the DFmax/rotor protocol. We also ob-
sealed all venous orifices (except for the inferior
tained corresponding regularity index (RI) maps (18).
vena cava) so as to control the level of intra-atrial
These maps provide a quantitative spectral analysis-
pressure. The intra-atrial pressure was maintained
based spatial distribution of signal regularity and
at 0-5 cm H 2O for the DFmax/rotor and HWD/CFAEs
fractionation (15). A realistic representation of optical
protocols, and at 12 cm H2 O for the ACh protocol.
signal-derived electrograms was also made possible
Then, AF was induced by burst pacing (10 Hz) and
by generating single-pixel ECGs at each pixel as pre-
remained
viously (18). Phase videos and (X,Y) coordinates of
sustained
for
the
duration
of
the
experiment.
the phase singularities or wavebreaks (WBs) locaSEE PAGE 270
tions were obtained after Hilbert transformation
of the fluorescent signal (15). Thus, WB (X,Y)
coordinates are superimposable to phase singularity
SIMULTANEOUS ENDOCARDIAL AND EPICARDIAL
(X,Y) coordinates generated by the separation of a
MAPPING. The optical mapping setup includes 2
wavefront into daughter wavelets. Then, WB density
charged coupled device cameras. One camera was
maps were constructed by color coding each pixel
connected to a cardio-endoscope, while the other
according to the number of WBs/pixel/s. The HWD
served to map the corresponding epicardial view as
regions were defined as an individualized group of
previously (14,16). As shown in Figure 1, the endo-
pixels with a WB density >30% than surrounding
scope was directed either towards the PLA after
pixels. HWD/CFAEs sites’ (X,Y) coordinates were then
introducing its tip through a minimal left ventricular
noted as reference for the positioning of the ablation
opening (Figure 1A) or toward the left atrial free
catheter tip during the HWD/CFAEs protocol. After the
wall (LAFW)—here the tip was advanced through
experiments, AF wave patterns were classified as
the superior vena cava and the inter-atrial septum
described in the Online Appendix. When patterns
(Figure 1B). The endocardial PLA view included
analyzed were rotors on both the epicardial and
the PV ostia and a portion of the left atrial roof
endocardial sides, it was possible to extrapolate the
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 1 Simultaneous Endocardial and Epicardial Mapping Setup: PLA and LAFW Views
Two CCD cameras, one of them connected to a deflectable endoscope were used to record from matching endocardial and epicardial views
of the PLA or LAFW regions. (A) PLA view. The endoscope was inserted in the left atrium (LA) through a minimal left ventricular (LV) opening
and across the mitral valve (MV). (B) LAFW view. Here, the endoscope was introduced in the LA through the inter-atrial septum. Ant. ¼
anterior; CCD ¼ charged couple device; LAFW ¼ left atrial free wall; LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein;
PLA ¼ posterior left atrium; Post. ¼ posterior; RA ¼ right atrium; RPV ¼ right pulmonary vein; RV ¼ right ventricle.
scroll wave’s filament meandering trajectory in rela-
located in the vicinity of DFmax regions, and not WB
tion with the detailed atrial anatomy (17).
regions distant from the DFmax region determined as
ABLATION PROTOCOLS. After AF initiation, DF and
HWD/CFAE maps were constructed after each 5-s
video recording. Then, a 4-F radio-frequency ablation catheter tip was positioned either on the endocardium—after
introducing
it
in
the
endoscope
working channel as previously (16)—or on the epicardium. For the DFmax/rotor protocol, a multiplerotation rotor or a figure-of-8 pattern located within
the DFmax domain was targeted. Precisely, the rotor
core or a region equidistant from the figure-of-8’s
detailed above. In both protocols, 1 to 6 ablation
points/heart (5 to 15 W, 30 to 60 s, Radionics, Inc.)
were performed. Then, post-ablation optical recordings were obtained to examine the impact of point
ablation on AF dynamics. The endpoints were merely
to deliver ablation in the DFmax/rotor region or in the
HDW region so as to enable optical video recordings
afterwards. In contrast with clinical ablation, our
endpoints were not AF termination or the prolongation of the AF cycle length.
cores was targeted. When rotors were only transient,
ACh PROTOCOL. In other isolated hearts (n ¼ 12), we
the center of the DFmax domain was localized and
increased ACh concentration in two distinct fashions:
targeted. For the HWD/CFAE protocol, the ablation
from stretch-related AF to 0.05 m mol/l ACh (n ¼ 6),
catheter tip was positioned on HWD/CFAE sites
and from stretch-related AF to 0.1 m mol/l ACh (n ¼ 6).
259
260
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 2 Ablation Protocols
ACh ¼ acetylcholine; AF ¼ atrial fibrillation; CFAE ¼ complex fractionated atrial electrogram; DFmax ¼ highest dominant frequency; HWD ¼
high wavebreak density; RF ¼ radio-frequency.
Simultaneous endocardial-epicardial optical videos
Figures 3A to 3C and 4A to 4C show the difference be-
were recorded before and after increasing ACh con-
tween the DFmax and the HWD/CFAE regions. As
centration as described above.
STATISTICAL ANALYSIS. Group data were expressed
as mean SD. Statistical comparisons were made:
1) maximum DF before and after ablation were performed by a paired Student t test; 2) filament average
lifespans by unpaired Student t test; and 3) average
DFs from different regions in the same animal were
compared with a 2-way analysis of variance with the
Bonferroni post-hoc test. Differences were considered
significant when p < 0.05.
RESULTS
DF/ROTOR AND HWD/CFAE PROTOCOLS. Figure 2
shows the 2 ablation protocols that we followed.
The left panel of Figure 2 shows the DFmax/rotor
protocol. After having obtained a DF map, the DFmax
region was delineated. When a multiple-rotation
previously shown (15): 1) the DFmax region is adjacent
to the HWD region; 2) regions with high signal regularity are found in the center of the DFmax domain; and
3) the boundaries of the DFmax domain correspond to
regions of HWD and low signal regularity. In Figures 3B
and 4B, representative examples of single-pixel recordings from a region of high regularity—located
within the DFmax domain—show that single-pixel
electrograms were regular in amplitude, and presented with one main deflection per activation. In
comparison, single-pixel electrograms from a region of
low regularity—located at the boundary of the DFmax
domain—were irregular in amplitude, and presented
with several deflections per activation. It should be
noted, however, that CFAE spatial distribution does
not necessarily overlap with HWD distribution as
CFAEs are unspecific electrogram features.
AF DYNAMICS MODULATION AFTER DFmax/ROTOR AND
rotor or a figure-of-8 pattern was visualized within
HWD/CFAEs
the DFmax domain, the rotor core or a region equi-
shows the 2 examples of DFmax/rotor ablation that
ABLATION. A F
t e r m i n a t i o n . Figure
3
distant from the figure-of-8’s cores was targeted.
led to AF termination. In sheep 1 (Figures 3A to 3C,
When rotors were only transient, the center of the
Online Video 1), we visualized a stable figure-of-8
DFmax domain was instead localized and targeted
pattern located on the anterosuperior aspect of the
(1 to 5 points, 5 to 15 W, 30 to 60 s). The HWD/CFAEs
LAFW. Because this stable re-entry was located within
protocol (right panel) corresponded to a point abla-
the DFmax domain, we targeted this figure-of-8 with
tion of the regions harboring the HWD, which
our ablation catheter. As shown in Figure 3D, AF
were also the locations where the most fractionated
terminated approximately 3 min after ablation. In
electrograms were recorded. These regions were
Figures 3E to 3G we present the other experiment in
delineated after having constructed HWD maps as
which DFmax/rotor ablation led to AF termination.
described in the methods section. For example,
After guiding the endoscopic tip towards the LAFW
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 3 DFmax/Rotor Ablation and AF Termination
(A to D) Sheep 1. (E to G) Sheep 4. (A) DF and RI maps. Representative example of corresponding DF and RI maps (see Methods section).
(B) Single-pixel electrograms from the DFmax region (black star) and from the boundary of the DFmax region (black X). (C) Region
targeted (white circle, upper panel). The lower panel is a phase video snapshot. The targeted location is the region located between the
2 figure-of-8’s cores (white arrow) (see also Online Video 1). (D) Bipolar electrograms from multiple atrial locations show AF termination after
ablation. (E) Representative bipolar electrogram recordings at the LAFW, PVJ, and RAA with their corresponding DF spectrum. The DFmax
domain was located at the LAFW and encompassed the rotor (not shown). (F) Epicardial-endocardial phase video snapshots showing a counterclockwise rotor. The ablation catheter tip was positioned on the rotor core (shaded region). Interestingly, this region is located at the junction
between thick and thin myocardium (photographic snapshot). (G) AF termination during DFmax/rotor ablation. AF ¼ atrial fibrillation; DF ¼
dominant frequency; PVJ ¼ pulmonary vein junction; RI ¼ regulatory index; RAA ¼ right atrial appendage; other abbreviations as in Figure 2.
261
262
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 4 High Wavebreak Density HWD/CFAEs Ablation
(A) Upper panel: RI map. In red: high regularity and low fractionation region; in blue: low regularity and high fractionation region.
Lower panel: HWD/CFAEs map (see Methods section). In red: HWD density region, in blue: low wavebreak density region. (B) Single-pixel
recordings in the high RI, low wavebreak density region corresponding to the DFmax domain (black star); and in the low RI, HWD/CFAEs region
corresponding to fractionated pseudo-electrograms (black X). (C) Upper panel: Corresponding DF map. Lower panel: Target HWD/CFAEs
sites (white circles). Abbreviations as in Figures 2 and 3.
DFmax domain, a counter-clockwise re-entrant ac-
are presented. In 5 of 7 animals in which DFmax/
tivity was visualized (Figure 3F). Then, the introduc-
rotor ablation did not lead to AF termination, the
tion of an ablation catheter into the endoscope
post-ablation DFmax domain remained in the vicinity
working channel enabled the targeting of the rotor
of the pre-ablation DFmax domain. Also, DFmax
core region (Figure 3F), and AF termination followed
values decreased significantly in the immediate post-
after 15 s (Figure 3E). In Figure 4, we present a
DFmax/rotor ablation. After the HWD/CFAE ablation,
representative example of HWD/CFAE ablation. The
however, the DFmax domain relocated from the
HWD map presented in Figures 4A and 4C depicts 2
LAFW to the PLA, and DFmax values did not change.
small regions with HWD, which were also the sites of
These results suggest that the on-target DFmax/rotor
fractionated electrograms (Figures 4B and 4C, white
ablation led the annihilation of the fibrillation-
circles and black X). These 2 regions were targeted
maintaining rotor in all hearts, but that in 5 of 7
with a total of 6 point-ablation attempts (3 epicardial
hearts, a secondary, much slower rotor succeeded in
and 3 endocardial). In this experiment, however, AF
maintaining the fibrillatory activity. In contrast, these
did not terminate (follow-up observation: 60 min).
results suggest that the nearly missed HWD/CFAE
Overall, after RF ablation of HWD/CFAE sites, none of
strategy spared the fibrillation-maintaining rotors
the hearts underwent AF termination (0 of 5 hearts,
and set the stage for large rotor drift from the LAFW
19 HWD/CFAE sites targeted). After DFmax/rotor
to the PV-PLA region.
ablation, however, AF terminated in 2 of 7 animals.
DFmax
Also, in the 5 of 7 hearts in which AF did not termi-
induced electrophysiological changes. As suggested by
nate, post-ablation AF dynamics were markedly
the results presented above, nearly missed point
different than before ablation.
ablation seems to act as an acute rotor unpinning
Post-ablation DF value and spatial distribution
event, followed by a large drift towards the PV-PLA
c h a n g e s . Figure 5 shows how the DFmax/rotor and
region. We designed another set of experiments to
the HWD/CFAE protocols differentially modulated AF
investigate whether acutely produced changes in
dynamics. We examined whether these ablation ap-
electrophysiological conditions—such as the ones
proaches led to DF value and DF distribution changes.
occurring after nearly missed ablation in the vicinity
In Figure 5, composite schematics of the DFmax
of the fibrillation-maintaining rotor (19,20)—may
domain location (upper panels), and a quantification
dislodge rotors and produce large rotor drifts. In
of pre- and post-ablation DFmax values (lower panel)
hearts in which AF had been initiated in the presence
domain
relocation upon pharmacologically
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 5 AF Dynamics Modulation After HWD/CFAEs and DFmax/Rotor Ablation
(Upper panels) Post-ablation DFmax domains relocation. The DFmax domains relocated from the LAFW to the PLA-PV region after HWD/CFAEs
(right panel) but not after DFmax/rotor ablation (left panel). (Lower panel) DFmax average values before and after HWD/CFAEs and
DFmax/rotor ablation. LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; PV ¼ pulmonary vein; RIPV ¼ right inferior
pulmonary vein; RSPV ¼ right superior pulmonary vein; other abbreviations as in Figures 1 to 3.
of an elevated intra-atrial pressure (stretch-related
LAFW region (Figure 6A). These results suggest
AF), we increased ACh concentration in 2 distinct
that large—but not moderate—electrophysiological
fashions: from stretch-related AF to 0.05 m mol/l ACh,
changes are sufficient for fibrillation sources to relo-
and from stretch-related AF to 0.1 mmol/l ACh. These
cate to a remote atrial region. Also, they are remi-
experimental conditions have been shown to enable
niscent of the relocation of AF drivers from the PV to
re-entry observation with optical mapping tech-
other atrial regions during persistent AF after intense
niques (11,18). Importantly under such conditions,
remodeling (4,9).
rotors locate either at the PV-PLA region or at the
Comparison
LAFW, depending on the ACh concentration. Thus,
p a t t e r n s a t t h e P V - P L A a n d a t t h e L A F W . The
we hypothesized that ACh concentration increased
above observation provided us an opportunity to
acutely in the same heart may lead to rotor relocation
examine 3-dimensional rotor—also known as scroll
of
3-dimensional
rotor
drifting
and drift. Figure 6A shows that during stretch-related
waves (SWs) (21)—dynamics in 2 distinct regions in
AF (in the absence of ACh), the DFmax domain was
the same heart. We iteratively focused the endoscopic
found at the PV-PLA region. A moderate ACh con-
tip and the corresponding epicardial camera on the
centration increase to 0.05 m mol/l led to a DF increase
DFmax regions before and after increasing ACh con-
at all locations, but the DFmax domain remained in
centration—for example, on the PLA during stretch-
the PV region (Figure 6A). In contrast, after a large
related AF, and on the LAFW at 0.1 m mol/l ACh.
increase in ACh concentration to 0.1 m mol/l, the
Under both conditions endocardial-epicardial rotors—
DFmax domain relocated from the PV-PLA to the
also known as I-filament SWs—were found (30% and
263
264
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 6 DFmax Location Before and After ACh Perfusion
(A) Average DF values before (n ¼ 12) and after increasing ACh concentration to either ACh 0.05 mmol/l (n ¼ 6) or 0.1 mmol/l (n ¼ 6).
(B) Average scroll wave filament lifespan. The inset is a schematic representation of the 3 main types of scroll waves that were detected at
ACh 0.1 mmol/l: I-, U-, and L-filament scroll waves. Ach ¼ acetylcholine; other abbreviations as in Figures 2 and 3.
53%, respectively, of all patterns of activation, see
drifted along the LAFW pectinate muscles. Alto-
Online Figure 1). Figure 6B further shows that at 0.1
gether, these results indicate that the LAFW is a re-
mmol/l ACh the average I-filament SW lifespan was
gion where rotors are prone to drift because of the
significantly longer than that of other SW pre-
many locations exhibiting a large myocardial thick-
sentations—named as L-type and U-type filament SWs
ness gradient. In comparison, in the PV-PLA region
(Online Appendix) (17,21). Figures 7A and 7B show 2
SWs remain relatively stable at the PV ostia, which are
consecutive
PLA locations with the largest myocardial thickness
endocardial-epicardial
phase
video
snapshots of the PV region superimposed with the
gradient.
corresponding high resolution still-picture of the
PLA. A counter-clockwise I-filament SW wherein
filament is anchored at the junction between the RPV
ostium and the central PLA is presented. Figure 7B
and Online Video 2 further shows that this SW only
drifted along the RPV ostial region. Figure 7C shows a
composite analysis of the I-filament SWs drifting
paths during stretch-related AF in 4 hearts with
distinct color dot time series. The I-filament SWs
DISCUSSION
In this work, we have examined the impact of 2 distinct
electrogram/mapping-based ablation strategies on AF
dynamics: on-target DFmax/rotor and nearly missed
HWD/CFAEs. Our main results are as follows:
1. On-target DFmax/rotor point ablation leads to
drifting paths were strikingly similar in that they
substantial decreases in AF fibrillatory frequency
remained within a very limited PLA-PV transitional
and to AF termination. In contrast, nearly missed
region. After ACh concentration was increased to 0.1
point ablation at HWD/CFAE regions—located in
mmol/l, however, Figure 8A and Online Video 3
the immediate vicinity of DFmax domains—does
describe entirely different SW drifting patterns.
not terminate AF and does not significantly
In Figure 8, endocardial-epicardial phase video
modulate fibrillation frequencies. Instead, nearly
snapshots superimposed with the corresponding
missed HWD/CFAE ablation causes DFmax do-
photographic snapshot of the LAFW are shown.
mains to relocate to the PV-PLA region.
A counter-clockwise I-filament SW is drifting along
2. Acute electrophysiological changes such as the
the pectinate muscle transition between thick and
ones occurring after a rapid ACh concentration in-
thin myocardium (Figure 8B). Figure 8C is a similar
crease, or after a nearly missed point ablation, set
analysis in 4 animals which shows that I-filament SWs
the stage for rotor relocation.
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 7 Scroll Wave Dynamics in the Absence of ACh (Stretch-Related AF)-PLA View
(A) Left panel: Schematic representation of the PLA view. Right panel: Two consecutive endocardial-epicardial phase video snapshots obtained
at the PLA were superimposed with their corresponding photographic background picture (also see Online Video 2). The filament trajectory is
extrapolated from simultaneous endocardial-epicardial views. The I-filament scroll wave is anchored at the junction between the pulmonary
vein and the PLA; black circles represent secondary wavebreaks. (B) Sequential meandering trajectory of the same I-filament scroll wave during
a 200 ms sequence (red to yellow). (C) In 4 hearts, 4 representative I-filaments scroll wave sequential meandering pattern. I-filaments
(red, orange, green, and blue points) are anchored at the junction between the pulmonary veins (thin) and the PLA (thick) myocardium.
Abbreviations as in Figures 1, 3, 5, and 6.
3. The LAFW is an atrial location prone to large rotor/
fastest and which harbor the highest DF should be
SW drifts along large myocardial thickness gradi-
chosen as targets (4,5,25,26). More recently, Narayan
ents. We show that SWs drift along the many
et al. (7) have shown the feasibility of localizing
pectinate muscles’ thin-thick myocardium transi-
fibrillatory
tions at the LAFW. In comparison, SW drift is
catheter-enabled electrogram phase analysis. Pre-
restricted to the PV ostia in the absence of ACh.
liminary results with this approach have shown substantial
reentrant
improvement
sources
of
the
with
a
ablation
64-spine
outcome
COMPARISON DFmax AND CFAE ELECTROGRAM/
(6,7). Regardless of the ablation strategy, however,
MAPPING-BASED ABLATION APPROACHES. Electrogram-
the mechanisms by which electrogram-based point
based AF ablation and the targeting of PV potentials
ablation may succeed or fail are unknown. For
have been described as the beacon of AF catheter
example, the difference in post-ablation dynamics
ablation (22,23). During the last 2 decades, in-
between targeting a rotor, the DFmax region, or re-
vestigators have proposed various strategies. Some
gions harboring CFAEs is unknown. Our previous
have targeted rapid, low amplitude and fractioned
work had shown that CFAEs mostly result from
electrograms (CFAEs) as a stand-alone (3), or as an
wavebreak formation at the boundaries of high fre-
adjuvant to anatomical PV isolation approaches (24).
quency domains such as the DFmax domain. In these
Others have suggested that regions activated the
regions, waves emanating from rotors experience
265
266
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
F I G U R E 8 Scroll Wave Dynamics at ACh 0.1
mmol/l-LAFW View
(A) Left-most panel: Schematic of the LAFW view. Right panels: Three consecutive endocardial-epicardial phase video snapshots at the LAFW
superimposed with their corresponding high resolution background picture (also see Online Video 3). An I-filament scroll wave meanders at
the junction between the thin and thick myocardial segment (black line), along the pectinate muscle border. (B) Sequential meandering
trajectory of the same I-filament scroll wave (red to yellow) during a 200 ms sequence. (C) In 4 hearts, 4 representative I-filaments scroll
waves sequential meandering and drifting trajectories. I-filaments (red, orange, green, and blue points) are meandering along pectinate
muscles at the junction between thin and thick myocardium. Abbreviations as in Figures 1 and 6.
beat-to-beat changes in directionality and local ve-
HWD/CFAE sites produce largely dissimilar post-
locity, and WBs occur (15). Importantly, the DFmax
ablation AF dynamics. On-target rotor ablation re-
domain, which encompasses high frequency rotors, is
sults in the termination of the fibrillation-maintaining
adjacent to the regions harboring HWD/CFAEs. These
rotor. This is followed by AF termination (2 of 7), or by
mechanistic observations could suggest that DFmax
the emergence of a much slower rotor (5 of 7). The fact
and rotor regions on the one hand, and HWD/CFAEs
that after DFmax/rotor ablation, AF either terminated
regions on the other hand, overlap; and that ablation
or evolved into a much slower atrial arrhythmia is
of one region is equivalent to aiming at the other. This
reminiscent of commonly described end-of-procedure
assumption could be seen as a unifying explanation for
atrial arrhythmias (27). By contrast, a nearly missed
the successes that the 3 aforementioned electrogram-
HWD/CFAE point ablation does not lead to AF termi-
based approaches have enjoyed (3,5,7). To the oppo-
nation and to post-ablation DFmax changes. Rather,
site, this work shows that small differences in the
nearly missed point ablation leads to rotor unpinning
targeting of the ablative energy result in entirely
and drift along the pectinate muscles’ large myocardial
distinct outcomes. Although the HWD/CFAE sites are
thickness gradients, or alternatively, to the formation
usually appended to the DFmax domain, we show here
of new rotors in remote atrial regions. Altogether,
that ablation at the DFmax/rotor and ablation of the
on-target
rotor
ablation
impinges
a
deadly
or
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
weakening blow to active fibrillatory sources. To the
ablation at the LAFW sets the stage for a large drift of
opposite, nearly missed rotor ablation seems to
fibrillation-maintaining rotors. This suggests that
strengthen fibrillatory sources in dislodging them from
nearly missed rotor ablation in patients with persis-
the ablated regions and in resetting AF dynamics.
tent AF and non-PV rotors (4,9) may make ablation
procedures more complex and lengthy.
ROLE OF LOCAL ELECTROPHYSIOLOGICAL CONDITIONS
IN PREFERENTIAL ROTORS/SWs ANCHORING, UNPINNING,
STUDY LIMITATIONS. We acknowledge that there
AND DRIFT. A 3-dimensional rotor—also known as an
are some inherent limitations to our experimental
SW—may be described as 2 rotors wherein the center
design. In particular, we recognize that comparing
of rotation is a filament. The atrium is a highly com-
the impact of ablation in 2 nearby region may not be
plex and heterogeneous anatomical region. The
directly relevant to currently implemented human
LAFW presents with an intricate web of pectinate
rotor ablation approaches. We fully agree, for
muscles—each of them is in essence a thickened atrial
example, that the proportion of AF episodes that
bundle connected to a thinner wall. By contrast,
terminated after either DFmax/rotor or HWD ap-
the PV-PLA region has only 2 regions with a large
proaches should not be compared side-by-side with
myocardial thickness gradient: the PV ostia and the
clinical outcomes. It is likely that, in the experiments,
septo-pulmonary bundle (28). In this work, we pre-
AF did not reproducibly terminate after DFmax/rotor
sent results showing that the drifting patterns of SWs
ablation because point ablation did not cover an area
at the LAFW are entirely distinct from the ones of SWs
sufficient to annihilate the dominant rotor. In com-
at the PV-PLA region. Previously, the scroll wave
parison, in the clinical setting, a relatively larger
filament tension was shown to determine where a
ablation area of a few cm 2 is routinely treated in the
scroll wave may drift and stabilize (29). When excit-
region of the rotor (6,7). Also, the mechanism by
ability is normal, the filament is submitted to a
which the post-HWD ablation DFmax regions relo-
“shrinking tension” and the SW tends to drift toward
cated to the PV region could alternatively be the
the thinnest myocardial regions. In contrast, in con-
termination of a leading rotor uncovering a second-
ditions of low excitability, the filament is submitted
ary PV rotor. We thus speculate that the termination
to a “growing tension” which results in the drifting
rate would have been higher if we had ablated
towards the thickest myocardium. Previously, we
beyond the “first” rotor, all subsequently mapped
have shown that atrial SWs are submitted to both
rotors. That these rotors were found in the PVs does
“growing and shrinking tension” in the atria and
not bear any weight in the debate on the need for
generally stabilize at the junction thin-thick myocar-
isolating PVs when rotors have been ablated. Finally,
dium (17). However, the manner in which the SW may
the perfusion of ACh as a means of modulating rotor
differentially anchor and drift after rapid changes in
dynamics is a mere approximation of the changes
electrophysiological conditions has not been experi-
that unfold after ablation; alternatively, the absence
mentally investigated. Here, our results indicate that
of interstitial fibrosis in normal isolated hearts could
the perfusion of ACh led to a drift and relocation of
explain why rotors may drift after HWD ablation.
SWs to a region characterized by an abundance of
Overall, we acknowledge that the optimal ablation
sites or sharp thickness gradients, for example, the
strategy and its precise extent (points, lines, or both)
thin-thick
LAFW
may not be extrapolated from this study. Whether
pectinate muscles (Figure 1). Our results further show
these findings are reproducible in the in vivo setting
that after having located at the LAFW, SWs are prone
and in humans will be the topic of subsequent
to large drifts (Figure 8B, Online Video 3). From a
investigations.
myocardium
transition
of
the
mechanistic point of view, we speculate that shorter
action potential duration (APD) after ACh perfusion
CONCLUSIONS
enables the drift of SW filaments along the pectinate
muscles of the LAFW.
Using high resolution optical mapping techniques,
After LAFW ablation, local changes in electro-
we compared post-ablation AF dynamics after 2
physiological properties such as depolarization and
electrogram-based ablation approaches: on-target
decreased excitability lead to filament destabilization.
DFmax/rotor or nearly missed HWD/CFAE point
These conditions are no longer favorable for SWs to
ablation. Although the HWD/CFAE and DFmax/rotor
remain at the LAFW and, as a consequence, SWs drift
regions are adjacent, our results show that on-target
towards the PV region where they stabilize (Figure 5).
DFmax/rotor point ablation is the most likely to halt
Thus, our results indicate that the anatomical speci-
rotors. In comparison, nearly missed HWD/CFAE does
ficities of the LAFW are such that nearly missed
not produce AF termination or frequency changes.
267
268
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
Atrial Fibrillation Dynamics
Rather, it likely leads to the unpinning and drifting
of rotors along large myocardial thickness gradients.
Overall, this experimental study provides a mechanistic framework to understand the difference between DFmax/rotor and CFAE ablation strategies.
These results suggest the need for a highly precise
localization of fibrillation sources, as well as for
obtaining information on the frequency of fibrillation
before targeting such sources. Potentially, they could
set the stage for designing improved ablation approaches so as to maximize AF termination and procedural success.
PERSPECTIVES
COMPETENCY IN MEDICAL KNOWLEDGE: AF
is the most common arrhythmia in adults. Catheter
ablation for AF has become a preferred alternative to
drug therapy for patients with drug-resistant AF. The
optimal ablation strategy remains a matter of debate
with recent emphasis put on electrogram-based strategies such as dominant frequency, complex fractionated electrograms, and rotor ablation.
TRANSLATIONAL OUTLOOK 1: Rotors are rapid
ACKNOWLEDGMENT The authors thank Jose Jalife
functionally reentrant sources which have been shown
for his support and fruitful discussions.
to play a key role in AF maintenance.
REPRINT REQUESTS AND CORRESPONDENCES TO:
Dr. Kalifa, Center for Arrhythmia Research, North
Campus Research Complex, University of Michigan,
2nd floor, Buildings 20 and 26, 2800 Plymouth Road,
Ann Arbor, Michigan 48109-2800. E-mail: kalifaj@
TRANSLATIONAL OUTLOOK 2: Rapid reentrant
sources generate both regular and fractionated electrograms. The former within the region activated at
the fastest frequency and the latter at the boundary
of high frequency domains.
umich.edu.
REFERENCES
1. Takahashi Y, O’Neill MD, Hocini M, et al. Effects
of stepwise ablation of chronic atrial fibrillation on
atrial electrical and mechanical properties. J Am
Coll Cardiol 2007;49:1306–14.
modulation) trial. J Am Coll Cardiol 2012;60:
628–36.
adrenocholinergic stimulation: interplay between
rotors and focal discharges. Heart Rhythm 2009;
6:1009–17.
2. O’Neill MD, Jais P, Takahashi Y, et al. The
stepwise ablation approach for chronic atrial
fibrillation—evidence for a cumulative effect.
J Interv Card Electrophysiol 2006;16:153–67.
8. Kumagai K, Sakamoto T, Nakamura K, et al.
Combined dominant frequency and complex fractionated atrial electrogram ablation after circumferential pulmonary vein isolation of atrial
fibrillation. J Cardiovasc Electrophysiol 2013;24:
975–83.
15. Kalifa J, Tanaka K, Zaitsev AV, et al.
Mechanisms of wave fractionation at boundaries
of high-frequency excitation in the posterior
left atrium of the isolated sheep heart
during atrial fibrillation. Circulation 2006;113:
3. Nademanee K, McKenzie J, Kosar E, et al. A new
approach for catheter ablation of atrial fibrillation:
9. Haissaguerre M, Hocini M, Denis A, et al. Driver
domains in persistent atrial fibrillation. Circulation
16. Kalifa J, Klos M, Zlochiver S, et al. Endoscopic
mapping of the electrophysiologic substrate. J Am
Coll Cardiol 2004;43:2044–53.
4. Sanders P, Berenfeld O, Hocini M, et al. Spectral
analysis identifies sites of high-frequency activity
maintaining atrial fibrillation in humans. Circulation 2005;112:789–97.
2014;130:530–8.
10. Mansour M, Mandapati R, Berenfeld O,
Chen J, Samie FH, Jalife J. Left-to-right gradient
of atrial frequencies during acute atrial fibrillation
in the isolated sheep heart. Circulation 2001;103:
2631–6.
626–33.
fluorescence mapping of the left atrium: a novel
experimental approach for high resolution endocardial mapping in the intact heart. Heart Rhythm
2007;4:916–24.
17. Yamazaki M, Mironov S, Taravant C, et al.
Heterogeneous atrial wall thickness and stretch
promote scroll waves anchoring during atrial
fibrillation. Cardiovasc Res 2012;94:48–57.
11. Berenfeld O, Mandapati R, Dixit S, et al.
Spatially distributed dominant excitation frequencies reveal hidden organization in atrial
18. Mandapati R, Skanes A, Chen J, Berenfeld O,
fibrillation in the Langendorff-perfused sheep
heart. J Cardiovasc Electrophysiol 2000;11:
869–79.
Jalife J. Stable microreentrant sources as a
mechanism of atrial fibrillation in the isolated
sheep heart. Circulation 2000;101:194–9.
6. Narayan SM, Krummen DE, Clopton P,
Shivkumar K, Miller JM. Direct or coincidental
elimination of stable rotors or focal sources may
explain successful atrial fibrillation ablation: ontreatment analysis of the confirm trial (conventional ablation for af with or without focal impulse
and rotor modulation). J Am Coll Cardiol 2013;62:
12. Kalifa J, Jalife J, Zaitsev AV, et al. Intra-atrial
pressure increases rate and organization of waves
emanating from the superior pulmonary veins
during atrial fibrillation. Circulation 2003;108:
668–71.
19. Nath S, DiMarco JP, Haines DE. Basic aspects
of radiofrequency catheter ablation. J Cardiovasc
Electrophysiol 1994;5:863–76.
138–47.
propagation into fibrillatory conduction across the
pectinate muscle network in the isolated sheep right
atrium. Circ Res 2002;90:1173–80.
5. Atienza F, Almendral J, Jalife J, et al. Real-time
dominant frequency mapping and ablation of
dominant frequency sites in atrial fibrillation with
left-to-right frequency gradients predicts longterm maintenance of sinus rhythm. Heart
Rhythm 2009;6:33–40.
7. Narayan SM, Krummen DE, Shivkumar K,
Clopton P, Rappel WJ, Miller JM. Treatment of
atrial fibrillation by the ablation of localized
sources: CONFIRM (conventional ablation for atrial
fibrillation with or without focal impulse and rotor
13. Berenfeld O, Zaitsev AV, Mironov SF, Pertsov AM,
Jalife J. Frequency-dependent breakdown of wave
14. Yamazaki M, Vaquero LM, Hou L, et al.
Mechanisms of stretch-induced atrial fibrillation in the presence and the absence of
20. Nath S, Lynch C 3rd, Whayne JG, Haines DE.
Cellular electrophysiological effects of hyperthermia on isolated guinea pig papillary muscle.
Implications for catheter ablation. Circulation
1993;88:1826–31.
21. Yamazaki M, Filgueiras-Rama D, Berenfeld O,
Kalifa J. Ectopic and reentrant activation patterns
in the posterior left atrium during stretch-related
atrial fibrillation. Prog Biophys Mol Biol 2012;
110:269–77.
Yamazaki et al.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. 1, NO. 4, 2015
AUGUST 2015:256–69
22. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic
beats originating in the pulmonary veins. N Engl J
Med 1998;339:659–66.
23. Jais P, Haissaguerre M, Shah DC, et al. A focal
source of atrial fibrillation treated by discrete radiofrequency ablation. Circulation 1997;95:572–6.
24. Oral H, Chugh A, Yoshida K, et al.
A randomized assessment of the incremental role
of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for
long-lasting persistent atrial fibrillation. J Am Coll
Cardiol 2009;53:782–9.
Atrial Fibrillation Dynamics
26. Lazar S, Dixit S, Marchlinski FE, Callans DJ,
Gerstenfeld EP. Presence of left-to-right atrial
frequency gradient in paroxysmal but not persistent atrial fibrillation in humans. Circulation 2004;
29. Wellner M, Berenfeld O, Jalife J, Pertsov AM.
Minimal principle for rotor filaments. Proc Natl
Acad Sci U S A 2002;99:8015–8.
110:3181–6.
27. Chae S, Oral H, Good E, et al. Atrial tachycardia after circumferential pulmonary vein
ablation of atrial fibrillation: mechanistic insights, results of catheter ablation, and risk
factors for recurrence. J Am Coll Cardiol 2007;
50:1781–7.
25. Jarman JW, Wong T, Kojodjojo P, et al.
Spatiotemporal behavior of high dominant frequency during paroxysmal and persistent atrial
28. Klos M, Calvo D, Yamazaki M, et al. Atrial
septopulmonary bundle of the posterior left
atrium provides a substrate for atrial fibrillation
initiation in a model of vagally mediated pulmonary vein tachycardia of the structurally normal
fibrillation in the human left atrium. Circ Arrhythm
Electrophysiol 2012;5:650–8.
heart. Circ
175–83.
Arrhythm Electrophysiol 2008;1:
KEY WORDS ablation, atrial fibrillation,
cholinergic stimulation, scroll wave
A PPE NDI X For accompanying videos
with analysis and legends as well as a supplemental figure and table, please see the online
version of this article.
Go to http://www.acc.org/jaccjournals-cme to take the CME
quiz for this article.
269