JACC: CARDIOVASCULAR IMAGING, VOL. 8, NO.

11, 2015

Letters to the Editor

NOVEMBER 2015:13406

thrombus was present in 6 of 10 SVGs at baseline versus

1 of 10 SVGs at 12 months (Figure 1C). Mean thrombus
area at baseline imaging was 0.37
0.28 mm2 . A
double-layer appearance was identied by OCT
imaging in the wall of all studied SVGs at the 12month follow-up. The median thickness of the inner
layer was 0.23 mm (interquartile range: 0.20 to 0.28
mm), and its area was 3.73 mm2 (interquartile range:
2.79 to 4.92 mm 2), representing 25% (interquartile
range: 22% to 29%) of the total vessel cross-sectional
area.
Three-dimensional (3D) SVG reconstruction was
performed using a previously validated methodology on the basis of the 3D luminal centerline
derived from angiographic projections (1,2). The obtained 3D lumen reconstructions were then used for
assessing the local endothelial shear stress (ESS) distribution by employing computational uid dynamics, as previously described (3). Baseline low ESS

Please note: Dr. Banerjee has served on the speakers bureaus for St. Jude
Medical, Medtronic, and Johnson & Johnson; has received research grants from
Boston Scientic; and has received consulting fees from Medtronic, Volcano,
and Merck. Dr. Brilakis has received consulting/speaker honoraria from Abbott
Vascular, Asahi, Boston Scientic, Elsevier, Somahlution, St. Jude Medical, and
Terumo; has received research support from Guerbet and InfraRedx; and his
spouse is an employee of Medtronic. All other authors have reported that they
have no relationships relevant to the contents of this paper to disclose.

REFERENCES
1. Bourantas CV, Papafaklis MI, Athanasiou L, et al. A new methodology for
accurate 3-dimensional coronary artery reconstruction using routine intravascular ultrasound and angiographic data: implications for widespread
assessment of endothelial shear stress in humans. EuroIntervention 2013;9:
58293.
2. Papafaklis MI, Bourantas CV, Yonetsu T, et al. Anatomically correct
3-dimensional coronary artery reconstruction using frequency domain optical
coherence tomographic and angiographic data: head-to-head comparison
with intravascular ultrasound for endothelial shear stress assessment in
humans. EuroIntervention 2015;11:40715.
3. Stone PH, Saito S, Takahashi S, et al. Prediction of progression of coronary
artery disease and clinical outcomes using vascular proling of endothelial
shear stress and arterial plaque characteristics: the PREDICTION Study. Circulation 2012;126:17281.

was associated with: 1) the largest decrease in lumen

area (p < 0.001 vs. moderate and high ESS); 2) the

3D Transthoracic Echocardiography Provides

largest increase in plaque burden (p 0.011 vs. high

Accurate Cross-Sectional Area of the RV

ESS); and 3) augmented brous neointimal area, as

Outow Tract

represented by the inner wall layer area at follow-up

(p 0.020 vs. moderate ESS and p 0.009 vs. high
ESS) (Figure 1D).
In summary, during the rst year after CABG, SVGs
undergo signicant lumen loss due to a combination
of wall thickening and negative remodeling. SVG
segments with the lowest ESS develop the largest
negative remodeling and neointima formation. These
ndings provide important insights into the pathogenesis of early SVG failure.
Anna P. Kotsia, MD
Michail I. Papafaklis, MD, PhD
Tesfaldet T. Michael, MD, MPH
Bavana V. Rangan, BDS, MPH
Matthias Peltz, MD
Michele Roesle, RN, BSN
Michael Jessen, MD
Bernice Willis, RN, BSN
Georgios Christopoulos, MD
Georgios Nakas, MD
Soa Giannitsi, MD
Lampros K. Michalis, MD
Dimitris I. Fotiadis, PhD
Subhash Banerjee, MD
Emmanouil S. Brilakis, MD, PhD*

Measurement of the cross-sectional area (CSA) of the

right ventricular outow tract (RVOT) has been
required to calculate stroke volume, especially for
estimation of the pulmonary blood ow/systemic
blood ow ratio. Current guidelines of the American
Society of Echocardiography state that the CSA should
be calculated using the RVOT diameter measured in
either the parasternal long-axis view or parasternal
short-axis view. Recently, the shape of the RVOT has
been reported to be oval when using 3-dimensional
(3D) transesophageal echocardiography (1); however,
the dynamic change of RVOT geometry in a cardiac
cycle

has

not

been

well

validated

with

any

modality. The aim of this study was to assess

morphological and dynamic features of the RVOT
using multidetector computed tomography (MDCT)
and to compare the CSA of the RVOT obtained by
MDCT and by 3D or 2-dimensional (2D) transthoracic
echocardiography (TTE).
A total of 20 patients with clinically indicated
contrast MDCT were prospectively enrolled between
May and September 2014 at Tokushima University
Hospital. This study was conducted in accordance
with the Declaration of Helsinki, was approved by the

*Dallas VA Medical Center (111A)

Institutional Review Board of the University of

4500 South Lancaster Road

Tokushima, and each subject gave written informed

Dallas, Texas 75216

consent. The MDCT examination was performed

E-mail: esbrilakis@gmail.com

using a 320-MDCT scanner (Aquilion ONE, Toshiba

http://dx.doi.org/10.1016/j.jcmg.2014.12.017

Medical

Downloaded From: http://imaging.onlinejacc.org/ by Kendra Marsh on 12/07/2015

Systems,

Tokyo,

Japan).

Retrospective

1343

1344

JACC: CARDIOVASCULAR IMAGING, VOL. 8, NO. 11, 2015

Letters to the Editor

NOVEMBER 2015:13406

F I G U R E 1 Correlations Between RVOT-CT and Echocardiography-Derived RVOT CSAs

The RVOT CSA obtained from 3-dimensional echocardiography (RVOT3D) correlated better with the RVOT-CT compared with those calculated
by RVOT long-axis diameter (RVOTlax), RVOT short-axis diameter (RVOTsax), and both (RVOToval).

electrocardiography-gated MDCT was performed; 10

the RVOT-CT (bias: 105.3 mm 2), and there was no

phases of images were reconstructed at a 10% R-R

signicant correlation between them (r 0.08,

interval. The RVOT was manually traced in the

p 0.75). The RVOTlax overestimated the RVOT-CT

111.3

mm 2),

and

there

was

modest

double-oblique transverse view, and the CSA (RVOT-

(bias:

CT) was measured at just below the pulmonary

correlation between them (r 0.59, p 0.01). The

annulus in all 10 phases (0% to 90% R-R interval).

RVOToval had a small bias (bias: 13.5 mm 2)

Echocardiographic studies were performed using the

compared with the RVOT-CT, but there was no

iE33 ultrasound imaging system (Philips Medical

signicant correlation between them (r 0.41, p

Systems, Andover, Massachusetts). We measured the

0.11). The correlation between the RVOT3D and the

long- and short-axis diameters of the RVOT in early

RVOT-CT was excellent (r 0.92, p < 0.001) with

systole by a 2D probe and calculated the CSA using

less bias (bias: 24.9 mm 2).

each diameter, assuming that the RVOT is circular

The present study is the rst to demonstrate the

(RVOTsax, RVOTlax) and using both diameters,

morphological and dynamic features of RVOT CSA by

assuming that the RVOT is oval (RVOToval). The

MDCT. The CSA of the RVOT calculated by the di-

RVOT CSA in the same cardiac phase was also

ameters in either view of 2D-TTE may be inaccurate in

measured ofine from the 3D dataset obtained by a 3D

some cases; the 3D-TTE provides more accurate

probe (RVOT3D). Three patients were excluded from

measurement of the parameter.

this analysis because of poor image quality.

The RVOT CSA changed dynamically in a cardiac
cycle. The pulmonary valve opened from 10% to 30%
phases, and the maximum CSA was dened in the
10% phase. Figure 1 shows comparisons among
various CSAs obtained by multimodalities (MDCT,
3D-TTE, and 2D-TTE). The RVOTsax underestimated

Downloaded From: http://imaging.onlinejacc.org/ by Kendra Marsh on 12/07/2015

Naoko Sawada, MD
Hirotsugu Yamada, MD, PhD*
Kenya Kusunose, MD, PhD
Shuji Hayashi, MD, PhD
Takashi Iwase, MD, PhD
Masataka Sata, MD, PhD

JACC: CARDIOVASCULAR IMAGING, VOL. 8, NO. 11, 2015

Letters to the Editor

NOVEMBER 2015:13406

*Department of Cardiovascular Medicine

Previously it was shown that statins improve LV

Tokushima University Hospital

longitudinal function and that was evaluated using

2-50-1 Kuramoto

different imaging techniques than speckle tracking

Tokushima 770-8503

(2). Mizuguchi et al. (3) reported that telmisartan

Japan

signicantly improved GLS in hypertensive patients,

nding

E-mail: yamadah@tokushima-u.ac.jp

but

http://dx.doi.org/10.1016/j.jcmg.2014.12.018

the question about the duration of therapy in the

Please note: The study protocol was approved by our institutional clinical
research ethics committee. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.

only

after

12

months.

This

raises

present study, which was not been presented. Perhaps

longer

duration

of

therapy

would

change

the

results. Furthermore, Motoki et al. (4) recently showed

that calcium-channel blockers also have favorable

REFERENCE
1. Izumo M, Shiota M, Saitoh T, et al. Non-circular shape of right ventricular
outow tract: a real-time 3-dimensional transesophageal echocardiography
study. Circ Cardiovasc Imaging 2012;5:6217.

inuence on GLS, even after only 3 months of therapy.

Beta-1 blockade, induced by bisoprolol, results in the
reduction of GLS (5). However, the percentage of the
T1DM patients who used beta-blockers in the present

Longitudinal Strain and Type 1 Diabetes

Mellitus: Are We on the Wrong Track?

study is not high enough to change the nal result,

especially in the subjects with normoalbuminuria.
Jensen et al. (1) showed that the usage of any of the
medications was not independently associated with

We have read with great interest the recently published

GLS. However, the present multivariate model consists

article by Jensen et al. (1) about the 2-dimensional

of some variables that interact (systolic and diastolic

mechanics in a large cohort of patients with type 1

blood pressure, body mass index and LV mass index).

diabetes mellitus (T1DM). The investigators reported

It would be helpful to perform a multivariate analysis

signicant difference in left ventricular (LV) global

considering the inuence of the variables from model

longitudinal strain (GLS) between the control subjects

3 on LV mass index and E/e0 in order to evaluate the

and the T1DM subjects with albuminuria, whereas the

effect of different medications on LV structure and

diabetic patients and those with normoalbuminuria

diastolic function in this population.

have similar GLS as the healthy control subjects do,

The Thousand & 1 Study raises many questions that

even though these patients have had T1DM without

need to be resolved and conrmed, but it certainly

tight glycemic control (glycosylated hemoglobin

represents a solid ground for further investigations.

>8%) for more than 2 decades.

We would like to discuss several important points
of this investigation. The absolute value of GLS in the

Marijana Tadic, MD, PhD*

Cesare Cuspidi, MD

control group is somewhat lower than expected for

*University

the healthy mid-age population. Interestingly, dia-

MisovicDedinje

stolic blood pressure and LV mass index are higher in

School of Medicine

the control group than in the T1DM patients, espe-

University of Belgrade

cially than in the T1DM subjects with normoalbumi-

Heroja Milana Tepica 1

nuria, probably due to medical therapy.

11000 Belgrade

The other important issue is the lack of association

Clinical

Hospital

Center

Dr.

Dragisa

Serbia

between the usage of various medications and LV

E-mail: marijana_tadic@hotmail.com

mechanical remodeling in T1DM. The investigators

http://dx.doi.org/10.1016/j.jcmg.2015.03.014

found that the treatment of statins, beta-blockers,

Please note: The authors have reported that they have no relationships relevant
to the contents of this paper to disclose.

calcium-channel

blockers,

angiotensin-converting

enzymes, and angiotensin II inhibitors or diuretics is

not associated with any improvement of GLS in the
diabetic population. The data on this topic are scarce
and inconsistent, and this kind of investigation with a
large population of uncomplicated diabetic patients
has been waiting for a long time. The previous investigations showed that all these drugs improve LV
structure (reducing LV mass) and systolic and diastolic
function. Is it possible that they are completely ineffective for improvement of LV longitudinal function?

REFERENCES
1. Jensen MT, Sogaard P, Andersen HU, et al. Global longitudinal strain is not
impaired in type 1 diabetes patients without albuminuria: the Thousand & 1
Study. J Am Coll Cardiol Img 2015;8:40010.
2. Mizuguchi Y, Oishi Y, Miyoshi H, Iuchi A, Nagase N, Oki T. Impact of statin
therapy on left ventricular function and carotid arterial stiffness in patients
with hypercholesterolemia. Circ J 2008;72:53844.
3. Mizuguchi Y, Oishi Y, Miyoshi H, Iuchi A, Nagase N, Oki T. Benecial effects
of telmisartan on left ventricular structure and function in patients with hypertension determined by two-dimensional strain imaging. J Hypertens 2009;
27:18929.

Downloaded From: http://imaging.onlinejacc.org/ by Kendra Marsh on 12/07/2015

1345