Live:Real Time Three-Dimensional Transesophageal Echocardiography in Percutaneous Closure of Atrial Septal Defects

Full text

RESEARCH FROM THE UNIVERSITY OF ALABAMA AT BIRMINGHAM
Live/Real Time Three-Dimensional Transesophageal
Echocardiography in Percutaneous Closure of Atrial
Septal Defects
Maneesha Bhaya, M.D.,* Ferit Onur Mutluer, M.D.,* Edward Mahan, M.D.,* Luke Mahan,* Ming C. Hsiung,
M.D.,†Wei–Hsian Yin, M.D., Ph.D.,†Jeng Wei, M.D., MsD,†Shen–Kou Tsai, M.D., Ph.D.,†Guang–Yu Zhao,
M.D.,†Wei–Hsian Yin, M.D.,‡Manish Pradhan, M.D.,* Rajesh Beniwal, M.D.,§ Deepak Joshi, M.D.,* Fatemeh
Nabavizadeh, M.D.,* Amitoj Singh, M.B.B.S.,* and Navin C. Nanda, M.D.*
*Division of Cardiovascular Disease, University of Alabama at Birmingham, Birmingham, Alabama; †Heart
Center, Division of Cardiology, Cheng Hsin General Hospital, Taipei, Taiwan; ‡Faculty of Medicine, National
Yang-Ming University, Taipei, Taiwan; and §S.S.R. Medical College, Belle Rive, Mauritius
Objectives: This study assessed the ability of live/real time three-dimensional transesophageal echocar-
diography (3DTEE) in measuring (1) atrial septal defect (ASD) maximum dimension, area, and adjacent
rim size, (2) ASD occluder left and right atrial disk size, (3) length of contact between the left atrial
(LA) disk and the aorta, and in (4) assessing device related complications such as residual shunt,
device embolization, and device encroachment upon adjacent cardiac structures. Materials and
Methods: 3DTEE images acquired during percutaneous ASD closure by the Amplatzer Septal Occluder
in 15 adult patients were retrospectively analyzed. Offline analysis was done using both the Philips
5500 ultrasound system and Philips QLAB software. 3D color flow Doppler images were used to assess
residual ASD shunting. Results: The Philips 5500 and Philips QLAB measurements correlated well for
ASD maximum dimension and area measurements. The Philips QLAB 3DTEE disk size measurements
also correlated well with the manufacturer obtained sizes. The aortic rim was deficient in 7 of the 15
patients, and the mean ASD occluder device size was 4 mm greater than the mean ASD maximum
dimension. The LA occluder disk was in contact with the aorta throughout the cardiac cycle in 12 of the
15 patients, and the LA occluder disk size correlated significantly with the contact length with the aorta.
Conclusion: Most of the patients demonstrated contact between the LA occluder disk and the aorta
throughout the cardiac cycle. 3DTEE may be useful in identifying patients at greater risk for aortic ero-
sion. (Echocardiography 2013;30:345-353)
Key words: atrial septal defect in adults, two-dimensional transesophageal echocardiography, live/real
time three-dimensional transesophageal echocardiography, percutaneous closure of atrial septal defect,
aortic erosion by atrial septal defect closure device
Percutaneous transvenous closure of secun-
dum atrial septal defects (ASDs) was first success-
fully performed in 1974,
1
but only in recent
years, the technique has become widespread. It
is considered a safe procedure with reported
complication rates from 6 to 11%.
2,3
The
Amplatzer Septal Occluder (ASO) is the most
commonly used closure device.
4
Two-dimensional transesophageal echocardi-
ography (2DTEE) provides superior images of the
interatrial septum compared with two-dimen-
sional transthoracic echocardiography (2DTTE).
5–7
2DTEE is recommended by the American Society
of Echocardiography (ASE) for percutaneous ASD
closure guidance and has been validated for the
measurement of ASD maximum dimension and
adjacent septal rims.
6,8–11
2DTEE remains the
most commonly used imaging modality world-
wide for obtaining these measurements.
Accurate measurement of ASD maximum
dimension and septal rim size is essential in
selecting the correct occluder device size. Under
sizing the device may lead to a residual shunt or
device embolization and oversizing may lead to a
residual shunt, impingement on adjacent cardiac
structures, device erosion, or cardiac perfora-
tion.
2,3,12
2DTEE has limitations in assessing ASD dimen-
sions and adjacent rim size. It generates thin slice
Dr. Maneesha Bhaya and Dr. Ferit Onur Mutluer worked
equally on the paper and are to be considered first authors.
Address for correspondence and reprint requests: Navin C.
Nanda, M.D., University of Alabama at Birmingham, Echo Lab
SW/S102, 619 19th Street South, Birmingham, AL 35249.
Fax: (205)-934-6747;
E-mail: nanda@uab.edu
345
©2013, Wiley Periodicals, Inc.
DOI: 10.1111/echo.12106 Echocardiography

views that are very user dependent in obtaining
optimum imaging positions and angles to prop-
erly visualize the ASD while maintaining a mental
three-dimensional reconstruction of the image
which may preclude comprehensive periproce-
dural and postprocedural ASD assessment.
13
3DTEE has the potential to circumvent the
limitations of 2DTEE. 3DTEE generates unique en
face ASD images allowing (1) reliable measure-
ments of the ASD maximum dimension and area
and (2) en face imaging of the implanted device.
The occluder left atrial (LA) and right atrial (RA)
disk position and size can then be accurately
assessed.
14
3DTEE can be easily performed in the cathe-
terization laboratory unlike 3DTTE, magnetic
resonance imaging, and computerized tomogra-
phy. Intracardiac echocardiography requires an
additional intravenous line and has the same limi-
tations inherent in 2DTTE. Fluoroscopic balloon
sizing produces only 2D images, increases proce-
dure time, and does not allow visualization of the
ASD rim.
There are reports of the utility of 3D transeso-
phageal echocardiographic reconstruction in
measuring the ASD maximum dimension and
rim size and of LA disks causing indentation of
the aorta.
15–18
No study has evaluated 3DTEE in
(1) measuring the length of contact (mm)
between the LA disk and aorta after device
deployment or (2) measurement of the LA and
RA occluder disk diameters. These parameters are
especially relevant in view of recent reports of
aortic erosion by the ASO device.
19,20
We therefore retrospectively analyzed 3DTEE
images in 15 patients who underwent percutane-
ous device closure of secundum ASDs. The
purpose of this study was to assess the ability of
3DTEE in measuring (1) ASD maximum dimension
and area, (2) adjacent ASD rim size, (3) LA and RA
occluder disk diameters, (4) length of contact
(mm) between LA disk and the aorta, and (5)
assessment of device-related complications such
as residual shunt, device embolization, and device
encroachment on adjacent cardiac structures.
Material and Methods:
The study group included 15 adult patients diag-
nosed with a secundum ASD by 2DTTE and
2DTEE. There were 13 females and 2 males rang-
ing in age from 17 to 73 years (mean
41.4 years). Fourteen of the 15 patients had a
single defect and 1 patient had multiple ASDs (4
large and several small fenestrated ASDs). One
patient with a single septal defect also had a
large interatrial septal aneurysm. Patients were
studied by 2DTEE and 3DTEE before, during and
after the occluder device placement, although
preprocedure 2D- and 3DTEE images were not
available in 1 patient. Preprocedural 2DTEE and
3DTEE color Doppler flow images were available
in 14 patients and 1 patient, respectively. Post-
procedural 2DTEE and 3DTEE color Doppler flow
images were available in only 1 patient. A Philips
IE-33 ultrasound system (Andover, MA, USA)
with a 2DTEE 2–7 MHz transducer and a 3DTEE
97-2t transducer was used in all patients and 3D
images were obtained as described previously
from our laboratory.
21
All patients received an
ASO (AGA Medical Corp., Golden Valley, MN,
USA) device and the patient with multiple septal
defects received 2 devices (1 ASO and 1 Amplat-
zer Septal Fenestrated Occluder). The 2DTEE ASD
maximum dimension using multiple planes and
confirmed by shunt visualization by color Dopp-
ler were measured. Sizing balloons were used in
2 patients. The operator(s) based device size
selection on 2DTEE measurements and the avail-
ability of devices in the laboratory at the time of
ASD closure. Dimensions were remeasured
retrospectively.
Measurement of the ASD Area and Maximum
Dimension:
B-Mode 3D datasets were cropped to view the
ASD en face from the left atrium (Fig. 1, movie clip
1), transferred to a Philips 5500 ultrasound system
and then measured in the non-multiplanar ren-
dering (MPR) mode as described previously.
15,22
The dataset with the clearest en face ASD view
from the LA aspect was again selected for analysis
using the MPR mode of the Philips Q-Lab 7.0
software as described in previous studies.
13,23
Three orthogonal planes (postero-anterior/green,
lateral/red, and transverse/blue) occupied 3
display screen quadrants and the fourth or lower
right quadrant contained the cropped, non-MPR
(full volume) en face ASD view. Either the
postero-anterior or the lateral plane was aligned
parallel to the long axis of the defect in the
non-MPR mode and then adjusted to view the
maximum inner dimension of the defect in the
cardiac cycle using images with minimal signal
dropout. The transverse plane was aligned paral-
lel to the maximum dimension of the ASD in the
MPR mode and the en face image of the ASD
was visualized in the corresponding window. The
ASD area was measured by planimetry avoiding
tissue dropout.
Measurement of the Septal Rims:
The 3D full-volume datasets were cropped to
identify the aortic valve and aorta utilizing the
clearest images. The aortic rim was the minimum
distance from the inner edge of the ASD to
the aortic outer wall and was assessed using the
Philips 5500 system non-MPR mode because the
Q-LAB MPR mode did not clearly display the
346
Bhaya, et al.

aortic rim due to image signal dropout. Septal
rims <5 mm were termed deficient as reported
previously.
17
An attempt was also made to mea-
sure the superior vena caval (SVC), inferior vena
caval (IVC), and mitral and posterior rims in the
same way as described above.
Measurement of LA and RA Occluder Disk
Size:
3DTEE datasets were obtained following device
deployment by centering the devices in the field
of view to image the disks in their entirety. Crop-
ping was done by first visualizing en face the LA
disk with its centrally attached central marker
band and then the RA disk with its centrally
located end screw as described previously from
our laboratory.
14
The dataset with the clearest
disk images were selected for further analysis in
the MPR mode.
Both the postero-anterior and the lateral
planes were moved parallel to the diameter of
the disk and perpendicular to each other in the
non-MPR image displayed in the right lower
quadrant. Both planes were adjusted to pass
through the central marker band of the LA disk
and the end screw of the RA disk. Small adjust-
ments enabled clear visualization of the disk
edges and then the disk radii were measured (the
distance from the disk edge to the middle of the
marker band and the end screw for the left and
RA disks, respectively) (Fig. 2, movie clip 2). The
diameter of each disk was obtained by doubling
the radius.
Measurement of the Length of Contact
Between the Left Atrial Disk and the Aorta:
The portion of the disk in contact with the aorta
without intervening septal tissue and the dura-
tion of contact relative to the cardiac cycle was
measured using the MPR mode. Encroachment
or contact of either disk with other cardiac struc-
tures such as the SVC/IVC, mitral, or tricuspid
valves was also evaluated.
Disk measurements following device deploy-
ment were easily obtained in the MPR mode
because the echogenic disks were well visualized
and the echo dropout of the disks was not signifi-
cant. As the disks were circular, the radius could
be measured if the central marker band or end
screw and a well-defined disk margin could be
visualized. Hence, it was not necessary to obtain
these measurements using the Philips 5500 ultra-
sound system non-MPR mode.
3DTEE Assessment of Device Complications:
Residual defects on B-mode 3DTEE were noted
following device deployment. Two patients had
color Doppler flow studies to evaluate the
Figure 1. Live/real time three-dimensional transesophageal echocardiography (Case #14). Full-volume image. Arrowhead points
to a large atrial septal defect. The aortic (AO) rim, shown by asterisks, measured 2 mm, the inferior vena cava (IVC) rim 11.6 mm
and the superior vena cava (SVC) rim 8.8 mm. RA =right atrium. (movie clip 1).
347
3DTEE in ASD Closure

presence of shunt signals through or around the
devices.
Intra-observer and Inter-observer Variability:
3DTEE measurements for the maximum dimen-
sion of ASD using QLAB, aortic rim using Philips
5500, SVC rim using QLAB, LA and RA disk size
and the contact length between the LA disk and
the aorta were repeated by the same observer
7 days later in a blinded fashion. The measure-
ments were also repeated by a second indepen-
dent observer blinded to the results of the first
observer and data from the manufacturer. Intra-
observer and inter-observer variability was
assessed using correlation coefficients and kappa
values.
Correlation Coefficients:
These were calculated for (a) 2DTEE measured
ASD maximum dimension and device size; (b)
3DTEE measured ASD maximum dimension and
device size; (c) ASD area and maximum dimen-
sion using the Philips 5500 ultrasound system
and QLAB; (d) SVC rim using the Philips 5500
Figure 2. Multiplanar Rendering Mode: In upper panels, top arrow arrow heads point to the left atrial (LA) disk, lower arrow-
heads to the right atrial (RA) disk. D1 represents the measurement of the radius of the LA disk from the middle of the central mar-
ker band to the outer edge of the disk. D2 represents the measurement of the radius of the RA disk from the middle of the end
screw to the outer edge of the disk. The arrow in the top left panel points to the area of contact of the LA disk with the aorta (AO).
Arrowhead in the right lower panel points to the LA disk covering the atrial septal defect. Movie clip 2 shows the LA disk in contact
with the aorta throughout the cardiac cycle.
348
Bhaya, et al.

ultrasound system and QLAB; (e) ASD area and
device waist area; (f) measured and manufac-
turer obtained LA and RA disk sizes; and (g)
contact length between the LA disk and the
aorta with LA disk size, device size, ASD maxi-
mum dimension, preprocedural aortic rim, and
ASD area. Statistical analysis was performed
using Medcalc (Mariakerke, Belgium) and
Microsoft (Redmond, WA, USA) Excel from the
Office 2010 software.
Results:
ASD Dimensions and Areas:
The ASD maximum dimensions measured by 2D-
TEE in all patients ranged from 3.8 to 32 mm
(mean 18.0 mm). ASD maximum dimensions
and areas measured in all patients by 3DTEE
using the Philips 5500 ultrasound system ranged
from 4.5 to 30 mm (mean 18.0 mm) and 13–
283 mm
2
(mean 149 mm
2
), respectively. The
ASD maximum dimension measured by 3DTEE
using the QLAB MPR mode in all patients ranged
from 7.2 to 35 mm (mean 19.4 mm). The area
measured in 12 of 14 patients by the QLAB MPR
mode ranged from 144 to 760 mm
2
(mean
400 mm
2
). It could not be measured in 2
patients because of tissue signal dropout.
Atrial septal defect maximum dimensions
measured by the Philips 5500 system and QLAB
correlated well with each other (r =0.99,
P<0.05). The correlation was lower for ASD area
(r =0.60; P <0.04). The mean ASD maximum
dimension measured by the Philips 5500 system
was equal to the 2DTEE measurement (18 mm,
range 5mmto+7 mm). The mean ASD maxi-
mum dimension measured by the QLAB MPR
mode was 1.4 mm greater than the 2DTEE mea-
surement (range: 5mmto+13 mm).
The sizes of the implanted devices ranged
from 8 to 36 mm (mean 23.3 mm). Both 2DTEE
and 3DTEE measured ASD maximum dimensions
correlated well with device size (r =0.84,
P<0.05; r =0.92, P <0.05, respectively)
(Figs. 3 and 4). 3DTEE measured ASD areas also
demonstrated good correlation with the calcu-
lated device waist area (r =0.80, P <0.05).
The single secundum ASDs did not conform
to a consistent geometric shape in the study
patients. One patient had multiple irregular
defects (4 large and several small fenestrated
defects).
Rim Sizes:
The aortic rim was visualized in 12 of 15 patients
and ranged from 2 to 6.1 mm (mean 4.1 mm)
using the Philips 5500 system. In the remaining 3
patients, the aortic rim was not visualized. The
aortic rim was classified as deficient (<5 mm) in 9
patients (Table I).
The SVC rim was visualized in 9 of 15 patients
and was measured using both the Philips
5500 system and the QLAB MPR mode and
ranged from 6 to 16 mm (mean 10.9 mm) and
8.7 to 14.6 mm (mean 11.1 mm), respectively
(Table I). Measurements obtained by both
methods correlated well with each other
(r =0.83, P <.05). The SVC rim was not visual-
ized in the remaining 6 patients and therefore
was not measured. The IVC rim was measured in
1 patient and was found adequate (11.6 mm). In
the remaining patients, the IVC was not visual-
ized. Mitral and posterior rims were not visualized
in any patient and therefore could not be mea-
sured. 3DTEE measured LA and RA disk sizes
using the QLAB MPR mode correlated well with
the manufacturer obtained sizes (r =0.99,
P<0.05; r =0.99, P <0.05, respectively)
(Figs. 5 and 6).
Figure 3. Correlation of atrial septal defect (ASD) maximum
dimension by two-dimensional transesophageal echocardiog-
raphy (2DTEE) with ASD device waist size obtained from the
manufacturer in 13 cases.
Figure 4. Correlation of atrial septal defect (ASD) maximum
dimension by live/real time three-dimensional transesopha-
geal echocardiography (3DTEE) using QLAB multiplanar ren-
dering (MPR) mode with ASD device waist size obtained from
the manufacturer in 13 cases.
349
3DTEE in ASD Closure

Length of Contact between the LA Disk and
Aorta:
The LA disk was in contact with the aorta in 12 of
15 patients and ranged from 1 to 7.4 mm (mean
3.7 mm). The device was in contact with the
aorta throughout the cardiac cycle in all of them.
In these cases, it moved with the aorta but the
motion was not exactly synchronous and the
points of contact of the disk with the aorta
slightly changed during the cardiac cycle (Fig. 2,
movie clip 2). In the remaining 3 patients (# 2, 7,
and 13 in Table I), the device was separated from
the aorta by an intervening septal rim which
measured 2.5, 3, and 6 mm, respectively. The
contact length with the aorta correlated well with
the LA disk size (r =0.66, P <0.05), device waist
size (r =0.59, P =0.02), and ASD maximum
dimension (r =0.60, P <0.05). There was poor
correlation between the contact length and pre-
procedural aortic rim and ASD area (r =0.35,
P=0.30, r =0.02, P =0.9 respectively).
Contact between the Device and Other
Adjacent Cardiac Structures:
The device was neither in contact with the wall
nor encroaching upon the lumen of the SVC in 9
patients in whom the SVC was imaged. The
device encroached upon the SVC lumen during
end-diastole in the patient with the aneurysmal
interatrial septum (Fig. 7, movie clip 3). The rela-
tion of the device to other cardiac structures such
as the mitral valve, IVC, and tricuspid valve was
not assessed as they were not visualized.
Device-Related Complications:
Residual shunts could not be evaluated in 14 of
15 patients studied because neither color Dopp-
ler flow nor bubble studies were recorded by 2D
and 3DTEE immediately following device implan-
tation. In the patient with an associated atrial
septal aneurysm, significant residual shunt was
noted by 3D color flow Doppler, and at 1 month
follow-up examination, the bubble study contin-
ued to demonstrate a residual shunt. In the
patient with multiple defects, the second device
(Amplatzer Septal Fenestrated Occluder) embol-
ized to the right iliac artery at the time of deploy-
ment of a third device. Surgery was performed to
retrieve the device and the ASD was closed. The
reason for embolization was not clear but may
have been related to mechanical dislodgement
by the delivery sheath at the time of deployment
of the third device. This patient has been
reported previously.
24
Figure 5. Correlation of left atrial (LA) disk size by live/real
time three-dimensional transesophageal echocardiography
(3DTEE) using QLAB multiplanar rendering (MPR) mode with
measurements from the manufacturer in 14 cases.
Figure 6. Correlation of right atrial (RA) disk size by live/real
time three-dimensional transesophageal echocardiography
(3DTEE) using QLAB multiplanar rendering (MPR) mode with
measurements from the manufacturer in 14 cases.
TABLE I
Aortic and Superior Vena Cava Rims before Atrial Septal
Defect Device Implantation and Length of Contact between
the Left Atrial Disk and the Aorta Following Device
Implantation
Case Ao rim (mm) SVC rim (mm) Contact (mm)
#1 4.2 12.7 4.3
#2 3.4 NM NC
#3 3.9 12.4 7.4
#4 NM NM 1
#5 3.9 NM 1
#6 3.7 NM 1.5
#7 3.5 NM NC
#8 NM 8.7 4.7
#9 6.1 14.6 5.6
#10 5.9 10.4 3
#11 5.3 12 4.8
#12 3.9 9 5.2
#13 NM NM NC
#14 2 8.8 2
#15 3.7 11 3.7
Ao =aortic; Contact =length of contact between left atrial
disk and aorta; NC =no contact; NM =not measured
because not visualized; SVC =superior vena cava.
350
Bhaya, et al.

Follow-up 2D and 3D transthoracic echocar-
diograms including color Doppler flow studies
performed 1–3 years after the procedure were
available in 10 of 15 (Table I. #2, 4, 5, 7, 8, 9,
10, 11, 14, and 15) study patients. None of these
patients had residual shunt or device related
complications.
Intra-observer and Inter-observer Variability:
The correlation coefficients for intra-observer and
inter-observer variability were as follows: (a) for
the ASD maximum dimension measured by
3DTEE (QLAB MPR): r >0.99, P <0.05 and
r>0.99, P <0.05, respectively. The kappa value
for inter-observer variability was 0.97; (b) aortic
rim: r 99, P <0.05 and r =0.94, P <0.05,
respectively. The kappa value for inter-observer
variability was 0.87; (c) SVC rim: r >0.99,
P<0.05 and r =0.63, P <0.05, respectively.
The kappa value for inter-observer variability was
0.58; (d) LA disk: r >0.99, P <0.05 and r =
0.98, P <0.05, respectively. The kappa value for
inter-observer variability was 0.93 (e) RA disk:
r>0.99, P <0.05 and r =0.97, P <0.05,
respectively. The kappa value for inter-observer
variability was 0.87, respectively, (f) contact
length between the LA disk and aorta: r >0.99,
P<0.05 and r =0.88, P <0.05, respectively.
The kappa value for inter-observer variability was
0.80.
Discussion:
The mathematical mean of ASD maximum
dimensions measured by 3DTEE and 2DTEE was
not statistically different in our study which is
similar to previous studies.
13,25
However, when
individual cases were considered, the differences
ranged from 5to+13 mm which could lead to
a wide variation in device size selection. Although
most authorities recommend use of a device
within 2 mm of the stretched balloon diame-
ter,
15,26,27
in our study the mean size of the
implanted device was 4 mm greater than the
mean maximum dimension of the ASD measured
by 3DTEE using the QLAB MPR mode. This over-
sizing may be partly attributed to the device size
availability at the time of the procedure. For
example, if the ASD size was 21 mm which
would require a 23 mm occluder and if it was
not available a 25 mm device would be used.
This would explain the higher correlation
obtained between the ASD maximum dimension
and the device waist diameters (which are both
linear dimensions) compared with the ASD area
and device waist area (squared linear values). This
oversizing mismatch of device size would be
expected to be mathematically amplified by
using the squared linear values and thus decrease
the area correlation values. The 3DTEE LA and RA
disk diameters correlated well with manufacturer
obtained measurements, substantiating the abil-
Figure 7. Live/real time three-dimensional transesophageal echocardiography (Case # 13). The arrowhead points to the right
atrial disk which is seen encroaching upon the lumen of the superior vena cava (SVC) (movie clip 7).
351
3DTEE in ASD Closure

ity of 3DTEE technology to obtain precise mea-
surements.
The ASD area and maximum dimensions mea-
sured using the thin slice QLAB MPR mode and
the thicker slice Philips 5500 ultrasound system
correlated well. However, echo signal dropout
was problematic with the QLAB MPR mode and
precluded measurement of the ASD area in 2
patients, including the patient with multiple
defects. Similarly, aortic rims could not be
measured using the QLAB MPR mode because of
tissue signal dropout and were measured using
thicker slices and the Philips 5500 ultrasound sys-
tem. In contrast, SVC rims could be reliably
measured using the QLAB MPR mode because of
minimal tissue signal dropouts in the 3D datasets
obtained by us. Our findings highlight the limita-
tion of QLAB MPR mode due to its thin sampling
slices and subsequent inability to reliably measure
curvilinear structures.
Atrial septal defect rims of at least 5 mm are
generally considered adequate for device
closure. ASD rims less than 5 mm are considered
deficient and may not be suitable for device
closure.
17,25
However, some operators consider
it safe to deploy devices in patients with small
aortic rims
10,28
; therefore, accurate rim size is
critical. The Food and Drug Administration
(FDA) premarket clinical trials of the ASO device
found that complications such as residual shunt-
ing or device embolization were minimized
when the selected device size was 1–2mm
greater than the 2DTEE and balloon-sized ASD
maximum dimension.
12
Some operators prefer
to oversize the occluder device by 3–4 mm espe-
cially in cases with deficient aortic rims.
29
One of the most interesting findings of our
study was that in 80% of the study patients,
the implanted LA disk was in contact with the
aorta throughout the cardiac cycle. Prior stud-
ies have also reported indentation of the aorta
by the LA disk but did not quantify the length
of contact between the LA disk and the aorta.
In this study, the ASD maximum dimension, LA
disk diameter, and the waist size of the
implanted device correlated significantly with
the length of contact between the LA disk and
the aorta. Aortic rim length in our study did
not correlate with the LA disk and the aorta
contact length, in contrast to one study that
reported that aortic rim length was the sole
statistically significant predictor of contact
between the LA disk and the aorta.
18
These findings are of considerable signifi-
cance given that in most of the reported ASO
device tissue erosions, erosion occurred at the
atrial dome near the aortic root. Most erosion
cases had a deficient aortic rim and/or over-
sized device.
19,20
Almost half of the current
study patients had deficient aortic rims (7 of
15) and the devices were considerably over-
sized. Aortic erosions from ASD closure device
have been reported to occur commonly within
72 hours but also as late as 6 years following
device implantation and have been associated
with serious sequelae such as cardiac tampon-
ade.
20,30,31
The incidence of aortic erosion has
been low but variable with DiBardino et al.
30
reporting an incidence of 0.28% and an FDA
pivotal study reporting no erosions.
31
These
results likely will be underestimates when
systematic long-term follow-up studies in this
country and abroad are reported. There has
been increasing concern recently regarding this
complication with the FDA convening an advi-
sory panel to address this issue. The panel
mandated tracking of these devices and closer
follow-up of these patients.
30
Also as men-
tioned before, the motion of the LA disk was
not exactly synchronous with the aortic motion
and this could result in friction and shear
strain. Thus, 3DTEE could aid in identifying
these higher risk patients that may need more
frequent postprocedure follow-up.
Limitations:
The main limitation of the study was the
inability of the QLAB software MPR mode thin
image slices in assessing the curvilinear (non-
planar) ASD adjacent rim size due to echo sig-
nal dropout. The thicker image sections of
the Philips 5500 non-MPR mode negated this
limitation.
Conclusions:
In the majority of the study patients, the LA disk
was in contact with the aorta throughout the
cardiac cycle, the occluder devices were over-
sized and almost half of the aortic rims were defi-
cient. The aortic contact length significantly
correlated with the LA disk size and the LA and
RA disk sizes measured by 3DTEE correlated well
with the manufacturer obtained sizes, substanti-
ating the ability of 3DTEE in obtaining precise
measurements. Routine use of 3DTEE may aid in
identifying patients at greater risk for aortic
erosion by the ASO device.
References
1. King TD, Mills NL: Nonoperative closure of atrial septal
defects. Surgery 1974;75:383–388.
2. Chessa M, Carminati M, Butera G, et al: Early and late
complications associated with transcatheter occlusion of
secundum atrial septal defect. J Am Coll Cardiol
2002;39:1061–1065.
3. Yared K, Baggish AL, Solis J, et al: Echocardiographic
assessment of percutaneous patent foramen ovale and
atrial septal defect closure complications. Circ Cardiovasc
Imaging 2009;2:141–149.
352
Bhaya, et al.

4. Masura J, Gavora P, Formanek A, et al: Transcatheter
closure of secundum atrial septal defects using the
new self-centering amplatzer septal occluder: Initial
human experience. Cathet Cardiovasc Diagn
1997;42:388–393.
5. Boccalandro F, Baptista E, Muench A, et al: Comparison
of intracardiac echocardiography versus transesophageal
echocardiography guidance for percutaneous transcath-
eter closure of atrial septal defect. Am J Cardiol
2004;93:437–440.
6. Butera G, Chessa M, Bossone E, et al: Transcatheter
closure of atrial septal defect under combined trans-
esophageal and intracardiac echocardiography. Echocar-
diography 2003;20:389–390.
7. Hijazi Z, Wang Z, Cao Q, et al: Transcatheter closure of
atrial septal defects and patent foramen ovale under
intracardiac echocardiographic guidance: Feasibility and
comparison with transesophageal echocardiography.
Catheter Cardiovasc Interv 2001;52:194–199.
8. Silvestry FE, Kerber RE, Brook MM, et al: Echocardiogra-
phy –Guided interventions. J Am Soc Echocardiogr
2009;22:213–231.
9. Vaidyanathan B, Simpson JM, Kumar RK: Transesopha-
geal echocardiography for device closure of atrial septal
defects. JACC Cardiovasc Imaging 2009;2:1238–1242.
10. Harper RW, Mottram PM, McGaw DJ: Closure of secun-
dum atrial septal effects with the Amplatzer septal occlu-
der device: Techniques and problems. Catheter
Cardiovasc Interv 2002;57:508–524.
11. Acar P, Saliba Z, Bonhoeffer P, et al: Influence of atrial
sepal defect anatomy in patient selection and assessment
of closure with the Cardioseal device. Eur Heart J
1999;21:573–581.
12. Du ZD, Cao QL, Rhodes J, et al: Choice of device size and
results of transcatheter closure of atrial septal defect
using the amplatzer septal occluder. J Interv Cardiol
2002;15:287–292.
13. Lodato JA, Cao QL, Weinert L, et al: Feasibility of real
time three-dimensional transesophageal for guidance of
percutaneous atrial septal defect closure. Eur J Echocardi-
ogr 2009;10:543–548.
14. Sinha A, Nanda NC, Misra V, et al: Live three dimensional
transthoracic echocardiographic assessment of transcath-
eter closure of atrial septal defect and patent foramen
ovale. Echocardiography 2004;8:749–753.
15. Chen FL, Hsuing MC, Hsieh KS, et al: Real time three-
dimensional transthoracic echocardiography for guiding
amplatzer septal occluder device deployment in patients
with atrial septal defect. Echocardiography 2006;23:763–
770.
16. Fuertes DG, Rubio DM, Ortiz MR, et al: Monitoring com-
plex secundum atrial septal defects percutaneous closure
with real time three-dimensional echocardiography.
Echocardiography 2012;29:729–734.
17. Roberson DA, Cui W, Patel D, et al: Three-dimensional
transesophageal echocardiography of atrial septal defect:
A qualitative and quantitative anatomic study. J Am Soc
Echocardiogr 2011;24:600–610.
18. Pepi M, Tamborini G, Bartorelli AL, et al: Usefulness of
three-dimensional echocardiographic reconstruction of
the amplatzer septal occluder in patients undergoing
atrial septal closure. Am J Cardiol 2004;94:1343–1347.
19. Amin Z, Hijazi ZM, Bass JL, et al: Erosion of amplatzer
septal occluder device after closure of secundum atrial
septal defects: Review of registry of complications and
recommendations to minimize future risk. Catheter Car-
diovasc Interv 2004;63:496–502.
20. Crawford GB, Brindis RG, Krucoff MW, et al: Percutane-
ous atrial septal occluder devices and cardiac erosion: A
review of the literature. Catheter Cardiovasc Interv
2012;80:157–167.
21. Singh P, Manda J, Hsiung MC, et al: Live/real time three-
dimensional transesophageal echocardiographic evalua-
tion of mitral and aortic valve prosthetic paravalvular
regurgitation. Echocardiography 2009;26:980–987.
22. Mehmood F, Vengala S, Nanda NC, et al: Usefulness of
live three-dimensional transthoracic echocardiography in
the characterization of atrial septal defects in adults.
Echocardiography 2004;21:707–713.
23. Skolnick A, Vavas E, Kronzon I, et al: Optimization of ASD
assessment using real time three-dimensional transesoph-
ageal echocardiography. Echocardiography 2009;26:233–
235.
24. Wei J, Hsuing MC, Tsai SK, et al: Atrial septal occluder
device embolization to an iliac artery: A case highlighting
the utility of three–dimensional transesophageal echocar-
diography during percutaneous closure. Echocardiogra-
phy 2012;29:1128–1131
25. Taniguchi M, Akagi T, Watanabe N, et al: Application of
real-time three-dimensional transesophageal echocardi-
ography using a matrix array probe for transcatheter
closure of atrial septal defect. J Am Soc Echocardiogr
2009;22:1114–1120.
26. Thanopoulos BD, Laskari CV, Tsaousis GS, et al: Closure
of atrial septal defects with the Amplatzer occlusion
device: Preliminary results. J Am Coll Cardiol 1998;31:
1110–1116.
27. Allen HD, Beekman RH III, Gardson A Jr, et al: Pediatric
therapeutic cardiac catheterization: A statement for
healthcare professionals from the Council on Cardiovas-
cular Disease in the Young. Circulation 1998;97:609–625.
28. Du ZD, Koenig P, Cao QL, et al: Comparison of trans-
catheter closure of secundum atrial septal defect using
the Amplatzer septal occluder associated with deficient
versus sufficient rims. Am J Cardiol 2002;90:865–869.
29. Demkow M, Ruzyllo W, Konka M, et al: Transvenous
closure of moderate and large secundum atrial septal
defects in adults using the Amplatzer septal occluder.
Catheter Cardiovasc Interv 2001;52:188–193.
30. DiBardino DJ, McElhinney DB, Kaza AK, et al: Analysis of
the US food and drug administration manufacturer and
user facility device experience database for adverse
events involving amplatzer septal occlude devices and
comparison with the society of thoracic surgery congeni-
tal cardiac surgery database. J Thorac Cardiovasc Surg
2009;137:1334–1341.
31. Diab K, Kenny D, Hijazi ZM: Erosions, erosions, and ero-
sions!Device closure of atrial septal defects: How safe is
safe? Catheter Cardiovasc Intgerv 2012;90:168–174.
Supporting Information
Additional Supporting Information may be found
in the online version of this article:
Movie clips for figures 1, 2, and 7.
353
3DTEE in ASD Closure