Content uploaded by Jun Endo
Author content
All content in this area was uploaded by Jun Endo on Feb 19, 2016
Content may be subject to copyright.
―84 ―
Tokai J Exp Clin Med., Vol. 33, No. 2, pp. 84-89, 2008
Visualization of Azygos Arch Valves using Computed Tomography:
Comparison of Scanning Delay Times
Jun ENDO, Tamaki ICHIKAWA, Jun KOIZUMI, Tomohiro YAMASHITA, *Ayako RO, Midori SAITO,
Yuri TANAKA, Kaoru ONOUE, Kazunobu HASHIDA, **Shu IKEDA, Yutaka IMAI
Department of Radiology, Tokai University School of Medicine
*Department of Legal medicine, Keio University
**Department of Radiological Technology, Tokai University School of Medicine
(Received March 7, 2008; Accepted April 22, 2008)
Objective: To evalu ate the frequenc y and appearance of azygos arch valves after short and long scanning
delays and high injection rates of contrast m aterial (CM) using a 64-slice multi-detector-row computed to-
mography (MDCT).
Methods: We retrospectively reviewed the findings from 264 contrast-en hanced MDCT chest examinations.
The rate of injection for 300 mg I/ml CM was 3.0 ml/sec; the short and long scanning delays were 20 and 180
sec, respectively. The presence of residual CM in the azygos arch valves and reflux of CM into the azygos
arch were recorded. A chi-squ are test was used to compare the frequency of residual CM in azygos arch
valves and reflux of CM into the azygos arch in both groups.
Results: Of the 132 examinations with short scan ning delays, 91 (68.9%) demonstrated residu al CM in azygos
arch valves and 103 (78.0%) demonstrated reflux of CM into the azygos arch. A significantly higher frequen-
cy of reflux of CMinto the azygos arch and residual CM in azygos arch valves was seen with short scanning
delays than with long sca nning delays (P < 0.05). However, no reflux of CM into azygos arch was seen with
long scanning delays.
Conclusions: Both reflux of CM into azygos arch valve and residual CM in the azygos arch were frequently
seen using short sca nning delays.
Key words: azygos arch, azygos arch valves, contrast-enha nced CT, contrast material
Jun ENDO, Department of Radiology, Tokai University School of Medicine, 143 Shimokasuya, Isehara, Kanagawa, 259 -1193, Japan Telephone number: Telephone number:Telephone number:
0463 -93-1121, Fax: 0463-93- 6827 E-mail address: jun@is.icc.u-tokai. ac.jp Fax: 0463-93- 6827 E -mail address: jun@is.icc.u-tokai.ac.jpFax: 0463-93- 6827 E -mail address: jun@is.icc.u-tokai.ac.jp E-mail address: jun@is.icc.u-tokai.ac.jpE-mail address: jun@is.icc.u-tokai.ac.jp
INTRODUCTION
Multi-detector-row computed tomography (MDCT)
is performed frequently using a high injection rate of
the contrast material (CM) to diagnose vascular lesions
[1-3]. The phenomena of reflux of CM into the azygos
arch from the superior vena cava (SVC) and residual
CM in the azygos arch valves are sometimes observed
after the high injection rate of the CM [4]. We evalu-
ated the frequency and appearance of the azygos arch
valves by conducting routine chest examinations and
injecting the CM at a low rate using single-slice helical
CT and 6-slice MDCT [5]. We reported that CM re-
mained more frequently in the azygos arch valves after
a rate of CM injection of 2 ml/sec using 6-slice MDCT
(45 of 100 examinations; 45%) than after a rate of
CM injection of 1 ml/second using single-slice helical
CT (15 of 100 examinations; 15%) [5]. Furthermore,
we evaluated the frequency and appearance of azygos
arch valves after high and low injection rates of CM
using 64-slice MDCT at contrast-enhanced CT of the
chest [6]. In our previous study, we showed that re-
sidual CM in azygos arch valves and reflux of CM into
the azygos arch were seen more frequently with a high
injection rate of CM with a short scanning delay than
with a low injection rate of CM with a long scanning
delay [5, 6].
The purpose of this study was to evaluate the fre-
quency and appearance of azygos arch valves using
64-slice MDCT after long and short scanning delays
using high injection rates of CM.
MATERIALS AND METHODS
Informed written consent was not required because
this was a retrospective study approved by our insti-
tutional review board. We conducted a computerized
search for all intravascular contrast-enhanced C T
examinations of the chest performed over a 16-month
period and identified 132 (15 men, 117 women) pa-
tients (mean age: 52.8 years; range: 24-88 years). CT
examinations were performed to rule out pulmonary
embolism and deep venous thrombosis. In almost all
the women, CT was performed before the gynecologic
surgery. Patients with heart failure, postoperative
thoracoplasty, pleuritis with excessive pleural effusion,
mediastinal tumor, lymphadenopathy, or massive
pulmonary embolism were excluded. The body weight
ranged from 40 to 75 kg. All patients were examined
in the supine position.
CT examinations were performed with a CM injec-
tion rate of 3.0 ml/sec and scanning delays of 25
(short) and 180 (long) sec. The short scanning delay
was performed to rule out pulmonary embolism, and
the long scanning delay was performed to evaluate
for deep venous thrombosis. All scans were performed
cranial-to-caudal from the apex of the lung. The es-
J. ENDO et al. /Visualization of Azygos Arch Valves on CT
―85 ―
timated scanning time at the level of the azygos arch
before and after administration of CM were 83 and
183 sec, respectively. All examinations were performed
by intravenously injecting isohexol (Omnipapue TM300;
Dai-ichi) at 2 ml/kg using a power injector. Fifty-
one patients were injected with CM in the right arm,
and another 51 patients were injected with CM in the
left arm. The examinations were performed using a
64-slice MDCT scanner (Somatom Cardiac Sensation
64; Siemens, Forchheim) with collimation of 0.6 mm,
0.5 sec of rotation time, and a reconstruction interval
of 5 mm.
Two board-certified radiologists (J. E. and T. I.) with
more than 15 years of experience interpreted the chest
CT images and reviewed all CT images using a picture
archiving and communication system workstation to
reach consensus. Images were systematically evaluated
for residual C M in azygos arch valves and reflux of
CM into the azygos arch from the SVC, which was
considered to be present when the enhancement of the
azygos arch and vein valves was more intense than
enhanced SVC (Fig. 1). The frequency of residual CM
in azygos arch valves and reflux of CM into the azygos
arch was recorded. When residual CM was seen in azy-
gos arch valves, the degree of residual CM in the azy-
gos arch valves was graded as Grade 1 if one side of
the valves was enhanced and as Grade 2 if both sides
of the valves were enhanced (Fig. 1). All axial CT im-
ages were evaluated on mediastinal windows (window
level: 50 Hounsfield units (HU); window width: 350
HU). However, setting of the window level and width
of the CT images were adjusted by the observer at the
time of image evaluation to minimize streak artifacts
and often approached those of bone window (window
level: 600 HU; window width: 2000 HU) (Fig. 1).
Chi-square analysis was used to compare the fre-
quency of residual CM in the azygos arch valves and
reflex of CM into the azygos arch with short and long
scanning delays. Mann–Whitney’s U test was used to
evaluate the correlation between injection site and fre-
quency of residual CM in azygos arch valves and CM
reflux into the azygos arch. Statistical significance was
set at P < 0.05.
RESULTS
Of the 264 examinations, 129 (49.0%) demonstrated
residual CM in the azygos arch valves, and 103 (39.0%)
demonstrated reflux of CM into the azygos arch. In a
short scanning delay, 91 (68.9%) demonstrated residual
CM in the azygos arch valves, and 103 (78.0%) demon-
strated reflux of CM into the azygos arch. In contrast,
during a long scanning delay, 38 (28.8%) demon-
strated residual CM in azygos arch valves and reflux
of CM was not observed in any of the examinations. A
significantly higher frequency of residual CM in azy-
gos arch valves was seen on short scanning delay than
on long scanning delay (P < 0.0001). The frequency
of residual CM in azygos arch valves and reflux of
CM into the azygos arch is shown in Table 1. Residual
CM in the azygos arch valves was demonstrated more
frequently when C M was administered in the right
arm (65 of 84 examinations; 78.3%) than in the left
arm (26 of 48 examinations; 54.1%) on short scanning
delay (P = 0.0057); whereas, the frequency of residual
CM in the azygos arch valves was independent of CM
injection site on long scanning delay (P = 0.7446). In
addition, the frequency of reflux of CM into the azy-
gos arch valves was independent of CM injection site
on short scanning delay (P = 0.1311). The degree of
residual CM in azygos arch valves is shown in Table 2.
Of the 129 examinations that showed residual CM in
azygos arch valves, 77 (60.2%) showed enhancement of
both sides of the azygos arch valve.
None of the patients had azygos lobe. One case of
Fig. 1. Reflux of contrast material into the azygos arch and residual contrast material in the azygos arch valves
a, b : Contrast-enhanced axial CT images on short scanning delay of bone window shows reflux of contrast mate-
rial from superior vena cava into the azygos arch (black arrow). Note residual contrast material in both sides of the
azygos arch valves (white arrows).
c: Contrast-enhanced axial CT image on long scanning delay of bone window shows
residual contrast material in both sides of the azygos arch valves (white arrows).
a b c
J. ENDO et al. /Visualization of Azygos Arch Valves on CT
―86 ―
aberrant left brachiocephalic vein with a high level of
entry to the SVC was observed when the CM was in-
jected in the left arm, and CM reflux into azygos arch
and residual CM in both sides of the azygos arch valve
were observed on short scanning delay (Fig. 2). There
was one case of persistence of left SVC with right arm
injection of the CM, and CM reflux into the azygos
arch and residual CM in one side of the azygos arch
valve were observed on short scanning delay.
DISCUSSION
The azygos vein is a large vein with a bicuspid valve
in the thoracic cavity (Fig. 3). The azygos system is a
paired venous pathway of the posterior thorax that
may be affected by numerous congenital and acquired
conditions. Contrast- enhanced CT is able to depict
anatomical changes in the azygos vein and disorders
of the azygos system (Fig. 2) [7-9]. A few articles
have reported reflux of CM into the azygos vein and
artifacts to be caused by CM in the azygos vein [10,
11]. In addition, previous case reports have suggested
that refulx of CM into the azygos vein may be an
indicator of abnormal right heart hemodynamics, such
as cardiac tamponade [12, 13]. However, Yeh’s group
has supported the stance that limited azygos reflux
is a non-specific phenomenon. Filling of the azygos
or hemiazygos veins after bolus injection of CM is
occasionally seen in healthy persons, and it is likely
explained by a simple gravitational effect [4].
However, no true functional valves exist in the
azygos system, and the presence of azygos arch valves
is not known to be of clinical importance [12]. Azygos
arch valves may result because of a ruptured vein
associated with the migration of the guidewire into
Table 2 Degree of residual contrast material in the azygos arch valves
Scanning delay Grade 1: one side Grade 2: both sides
Short (N=91) 24 67
Long (N= 38) 28 10
Total (N=129) 52 77
Table 1 Frequency of residual contrast material in the azygos arch valves and reflux of con-
trast material into the azygos arch
a. Residual CM in the azgos arch valves
Short scanning delay Long scanning delay
Right side injection of CM 78.3% (65/84) 29.8% (25/84)
Left side injection of CM 54.2% (26/48) 27.1% (13/48)
Total 68.9% (91/132) 28.8% (38/132)
b. Reflux of CM into the azygos arch on short scanning delay
Right side injection of CM 82.1% (69/84)
Left side injection of CM 70.8% (34/48)
Total 78.0% (103/132)
CM: contrast material
*
Fig. 2. 57-years-old women with the
aberrant left brachiocephalic
vein
a : Contrast-enhanced axial
CT ima ge on sh ort sc an-
ning delay with left side of
injection of contrast material
shows the aberrant left bra-
chiocephalic vein (asterisk)
courses behind the ascending
aorta and enters the superior
vena cave. Note res idual
contrast material in one side
of the azygos arch valves (ar-
row).
c: Coronal reformatted im-
age shows the aberrant left
brachiocephalic vein (arrow).
a
b
J. ENDO et al. /Visualization of Azygos Arch Valves on CT
―87 ―
azygos arch valves when a central venous catheter is
mistakenly inserted into the azygos arch. We found
that the frequency of CM reflux into the azygos arch
and residual CM in azygos arch valves were associated
with the injection rate, scanning delay time, injection
site, concentration of CM, and other factors [5, 6]. Yeh
and colleagues reported that a high frequency of CM
refluxed into the azygos vein in case of high injection
rates (3.5-5.0 ml/sec) as compared with the low injec-
tion rates (2.0-3.5 ml/sec) of highly concentrated CM
(350 mg I/ml) [4]. Many common limitations includ-
ing the use of various injection rates of CM, different
scanning delays, collimations, and thicknesses of the
reconstruction slice were seen in our previous study
and that conducted by Yeh and colleagues. As a result,
we were unable to precisely evaluate the frequency of
reflux of CM into the azygos arch and of residual CM
in azygos arch valves associated with scanning delay
[5, 6]. In the present study, we used the same injection
rate of CM and the same collimation and thickness
of the reconstruction slice to evaluate the frequency
of CM reflux into the azygos arch and residual CM
in azygos arch valves based on scanning delay. The
results showed that reflux of CM into azygos arch and
residual CM in azygos arch valves were seen more
frequently with short scanning delays than with long
scanning delays at a CM injection rate of 3.0 ml/sec.
The results of the present study agree with our previ-
ous study, i.e., that residual CM in azygos arch valves
was more frequently seen after right arm injection of
CM than after left arm injection of CM on short scan-
ning delay. Because residual CM in azygos arch valves
was washed out during long scanning delay, the injec-
tion site had no effect on the frequency of residual CM
in azygos arch valves. We previously reported that the
age of the patient was independent of the frequency of
residual CM in the azygos arch valves and CM reflux
into azygos arch [6].
When only a delayed scan was performed, the re-
flux of CM into the azygos arch subsided, which made
*
a b
Fig. 3. 57-years-old women with azygos lobe
a: Contrast-enhanced axial CT image shows azygos arch valves (arrows).
b: Axial CT of lung window shows the azygos lobe (asterisk).
the vein more inconspicuous. Consequently, residual
CM in the cusps of azygos arch valves tended to be
visualized as isolated areas of high density and could
easily be mistaken for other hyperdense structures,
such as calcified lymph nodes, surgical clips, tracheal
calcifications, or calcification of mediastinal tumor
(Fig. 4). The possibility of frequency of residual CM
in the azygos arch valves is increased when the azygos
arch valves are compressed by lung tumors or atelecta-
sis and by excessive pleural or pericardial effusion (Fig.
5). We had measured CT attenuations of residual CM
in the azygos arch valves, and those were more than
800 HU [5]. Image evaluation on bone window was
useful to distinguish residual CM in the azygos arch
valves from as calcified lymph nodes or surgical clips
(Fig. 4). It is important to understand the normal ap-
pearance and frequency of residual CM in the azygos
arch valves. We found that residual CM in one side of
the azygos arch valves was seen in 28 of 38 examina-
tions and that residual CM in both sides of the azygos
arch valves was seen in 10 of 38 examinations at a CM
injection rate of 3 ml/sec and with a long scanning
delay.
We obtained a huge volume of high-quality mul-
tiplanar reformation images with the 64-slice MDCT
[14]. The azygos vein valves were reported to be pres-
ent within 4 cm of the SVC, and the median distance
of the azygos arch valve from the SVC was 1.9 cm on
CT [4]. We found that the mean maximum length
of the azygos arch valves was 8.51±0.42 mm [5].
If residual CM is suspected on axial image, the prob-
lem can be solved by adding multiplanar reformatted,
three dimensional and maximum intensity projection
images (Fig. 5, 7).
High temporal and spatial resolution of MDC T
results in visualization of structures and pathologies,
which were undetected until date. The knowledge of
minute anatomy and structures, such as azygos arch
valves, is considered important in enhancing the diag-
nostic capabilities of MDCT.
*
J. ENDO et al. /Visualization of Azygos Arch Valves on CT
―88 ―
Fig. 4. a: Gross specimen of the azygos vein demonstrates a bicuspid valve of the azygos arch (arrows).
b: Schem of the aztgos vein and azygos arch valves (arrows).
b
a
*
Fig. 5. 58-years-old man with post-operative state
of esophageal cancer
Coronal reformatted image on bone win-
dow shows residual contrast material in
both sides of the azygos arch valves (arrows)
distinguished from multiple surgical clips
Fig. 6. 48-years-old man with lung cancer
Contrast-enhanced axial CT image shows residual contrast
material in one side of the azygos arch valves (arrow) adja-
cent to lung tumor and atelectasis (asterisk).
J. ENDO et al. /Visualization of Azygos Arch Valves on CT
―89 ―
In conclusion, our results indicate that residual CM
in the azygos arch valves and reflux of CM into the
azygos arch after high injection rates of CM were seen
more frequently with a short scanning delay than with
a long scanning delay. Residual CM in both sides of
azygos arch valves was frequently seen after a high
rate of CM injection with a short scanning delay.
ACKNOWLEDGEMENTS
The authors are grateful to Fumiko Kimura M. D.
who reviewed a part of this manuscript with her excel-
lent sugguestions.
REFERENCES
1) Schoellnast H, Deutschmann HA, Friz GA , Stessel U, Schffler
GJ, Tillich M (2005) MD CT angiography of the pulmonary
arteries: influence of iodine flow concentration on vessel attenu-
ation and visualization. Am J Roentgenol 184: 1935
2) Erdogan N, Akar N, Vural M, Vural M, Canbay A , Kayhan
T, Sahin D, Diker E, Aydogdu S. (20 06) Diagnostic value of
16- slice multidetedtor computed tomography in symptomatic
patients with suspected significant obstruction coronary artery
disease. Heart Vessels 21: 278-284
3) Hayashi H, Hidaka F, Kunazaki T, Ochi M (2006) Serial assess-
ment of the development of inflammatory abdominal aortic
aneurysm from ordi nary athero sclerotic abdominal aortuic
aneurysm using multidector-row computed tomographic angiog-
raphy. Heart Vessels 21: 334-337
4) Yeh BJ, Coakley FV, Sanchez HC, Wilson MW, Gotway MB
(20 04) Azygos arch valves: prevalence and app ea rance at
contrast-enhanced CT. Radiology 230: 111-115
5) Ichikawa T, Endo J, Koizumi J, Saito M, Kawada S, Myojin
K, Kobayashi M, Imai Y, Tsukune Y. (2006) Prevalence and
appearance of the azygos arch valves on contrast enhanced CT:
comparison of injection rates of contrast materialJpn. J. Med.
Imaging 25: 65-75 (Article in Japanese)
6) Ichikawa T, Endo J, Koizumi J, Ro A, Kobayashi, Saito M,
Kawada S, H ashimoto T, Imai Y, (20 07) Visualizat ion of
the Azygos A rch Valves on Multi-d etec tor-row Comp uted
Tomography. Heart Vessels 23: 118-123
7) Sm ath er s R L , B us chi AJ, Pop e T L Jr, Brenbr id gi AN ,
Williamson BR (1982) The azygos arch: normal and pathologic
CT appearance. Am J Roentgenol 139: 477-483
8) Dudiak CM, Olson MC, Posniak HV (1991) C T evaluation of
congenital and acquitted abnormaloities of the azygos system.
Radiographics 11: 233-246
9) Takasugi JE, Godwin JD (1989) CT appearance of the retroaor-
tic anastomases of the azygos system. Am J Roentgenol 154:
41-44
10) Godwin JD, Webb WR (1982) Contrast-related flow phenomena
mimicking pathology on thoracic computed tomogr aphy. J
Comput Assist Tomogr 6: 460 -464
11) Gotway MB, Patel R, Webb WR (2000) Helical CT for the evalu-
ation of suspected acute pulmonary embolim: diagnostic pitfalls.
J Comput Assist Tomogr 24: 267-273
12) Harries SR, Fox BM, Roobottom C A (1998) Azygos reflux: a CT
sign of cardiac tamponade. Clin Radiol 53: 702-704
13) Price R, Duell R, Wyatt S. (1999) Azygos reflux: a CT sign of
cardiac tamponade. Clin Radiol 54: 271-272
14) Nishino M, Kubo T, Kataoka ML, Gautam S, Raptopoulos V,
Hatabu H (2006) Evaluation of pulmonary embolisms using
coronal reformations on 64-row multidetector-row computed
tomography comparison with axial images. J Comput Assist
Tomogr 30: 233-237
*
a b
Fig. 7. Typical appearance of the azygos arch valves.
a: Maximum intensity projection image shows residual contrast material in both sides of tha azygos arch valves (black
arrow).
b: Volume rendering image shows the azygos arch valves (white arrow) posterior to the superior vena cave (asterisk).
