Michele Cimrelda
S. West, Cooper,
MD MD
#{149} Brian
Gallbladder in Patients
Index
terms:
Gallbladder,
1 #{149} Gallbladder,
Gallbladder, den, vanices,
US studies, 762.75 1991;
MD
CT studies, MR studies, 762.1214 762.1298 #{149} Gallblad-
T
the
systemic documenting
collateral portal
MATERIALS
AND
of Radiology,
Georgetown
is useful hyper-
in
Fifty consecutive
METHODS
patients
with
a diag-
nosis of portal hypertension based on clinical data (physical examination findings and/or free and wedged hepatic vein pressure measurements) initially exam-
med in our department during a 2-year period were included in our study. All CT examinations were performed with a CT/T 9800 scanner (GE Medical Milwaukee)
by using
a contrast
material-enhanced, dynamic rapid-sequence technique. Scanning parameters included a 2-second scan time, 120 kVp, 200 mA.
A volume
60% iodinated
From the 1989 RSNA scientific assembly. Received sion received November 13; accepted November ment of Radiology, The Johns Hopkins Hospital, © RSNA, 1991
vessels venous
tension and, in some cases, may be the first indication of its presence. Despite the wide variety of collateral pathways that have been described, relatively little attention has been focused on those that reside within the gallbladder wall and pericholecystic bed. The purpose of this study was to review the spectrum of imaging appearances of gallbladder varices in patients with portal vein hypertension.
System, #{149}
179:179-182
Department
S. Hayes,
DO.
MD
angiographic, sonogmaphic, and computed tomographic (CT) appearance of portal venous hypertension and portal vein thrombosis have been well documented in the literature (1-il). More recently, magnetic resonance (MR) imaging and duplex and colon Doppler flow ultrasound (US) have also been used for evaluation of the portal venous systern (12-17). Demonstration of porto-
modt,
From
#{149} Wendelin
HE
lamate
I
C. Honii, MD K. Zeman,
#{149} Robert
Varices: Imaging Findings with Portal Hypertension’
and Radiology
#{149} Steven
M. Silverman,
A retrospective review of the mcdical and imaging records of 50 patients with portal hypertension cxamined in the authors’ department during a 2-year period identified six patients with gallbladder wall vandes. Imaging studies performed in these patients included computed tomography (CT) (four patients), duplex and color Doppler flow (five patients), and magnetic resonance (MR) (four patients). Five of six patients with gallbladder vanices had portal vein thrombosis. Anechoic areas within the gallbladder wall detected with ultrasonography could be distinguished from intramural edema by using duplex on color Doppler flow imaging in all five patients in whom it was used. Contrast material enhancement of these vanices was detected with CT in three patients, two of whom also had adjacent mesentenic collaterals. Gradient-echo MR imaging (fast imaging in steady precession/fast low-angle shot) showed flow-nelated enhancement within the gallbladder wall in two patients. The presence of gallbladder wall vanices may imply the presence of portal vein thrombosis. Since these vanices can be a source of major blood loss, surgeons must be made aware of them when operating on patients with portal hypertension.
762.121
MD
S. Gamra,
#{149} Paul
contrast
meglumine,
St Louis)
University
of 100-150
material
Conray-60;
was
Medical
injected
Center,
mL of
(iotha-
Washington,
at a rate
DC.
of i.2-1.5
mL/sec.
Real-
time, duplex, and color flow sonography was perfomed with an Ubtramark 9 sonographic unit (Advanced Technology Laboratories, Bothell, Wash). MR images were acquired with a Magnetom imager (Siemens, Iselin, NJ) operating at 1.5 T. Spin-echo sequences with a repetition time (TR) of 400 msec and echo time (TE) of i5 msec (400/i5) (Ti) and 2,000/30, 80, 150 (triple-echo T2) were performed in the transaxial and coronal planes with a 256 )( 256 matrix and 8-mm-thick sections. Fast imaging in steady precession (FISP) flow-related sequences with a 45#{176} flip angle and 50/12 were also performed. In two patients, fast low-angle shot (FLASH) sequences with a flip angle of 30#{176} and 30/10 were also performed. Angiography was performed in two patients and
consisted
rior
and
debayed
of injections
inferior portal
into
the
supe-
artery with venous phase filming. were reviewed by three inmesentenic
images dependent observers. A consensus session was subsequently held, and the images were assigned to one of three groups depending on the ease of detection of the gallbladder varices: vanices not seen, vailces adequately seen, and vanices well seen (Table). The final groups were assigned on the basis of agreement between the three reviewers. All
RESULTS Of the 50 cases reviewed, six (12%) were found to have gallbladder wall varices. Although these six patients had documented portal hypertension, five were found to also have portal vein thrombosis. The diagnosis of portal vein thrombosis was made prospectively based on either the absence of flow in or cavernous transformation
Mallinck-
as a monopha-
June 12, 1990; revision requested July 16. Address reprint requests to M.S.W., 600 N Wolfe St. Baltimore, MD 21205.
sic bolus
the
flow-related
al vessels
in the
of the
portal
MR
images,
porta
vein
on
colbater-
hepatis
and
no
portal vein identified with contrastenhanced CT, or lack of flow in the portal vein detected with duplex so-
6; reviDepartAbbreviations:
TE
=
echo
time, TR
-
repeti-
tion time.
179
nogmaphy.
Thrombus
in the
extrahepatic
portal
vein
was portion
identified of the
patients and in both the intraand extrahepatic portions in three patients. Percutaneous liver biopsy results showed that all six patients had cirrhosis, suggesting that the portal vein thrombosis was not acute but secondary to portal yenous hypertension. In addition, all patients had prior evidence of gastrointestinal bleeding: Three had documented esophageal vanices, one had hemobilia, and two had an undetemmined source. The reasons for imaging were deteriorating liver function (three patients), routine follow-up (two patients), and right-upper-quadrant pain (one patient). In three patients the gallbladder vanices were surgically proved. Anechoic areas in the gallbladder wall showed venous flow within them at duplex or color Doppler flow imaging in four patients in whom it was used (Fig 1). In a fifth patient, only thickening of the gallbladder wall was detected on the two-dimensional US images; however, venous flow within the wall could be seen with colon Doppler flow (Fig 2). These anechoic areas appeared serpentine (in
four
in two
patients)
or as a single
demonstrable
venous
1.
flow
in
normal gallbladder wall. Contrast enhancement in the vances was present on CT scans in four of five patients, three of whom also demonstrated contact between adjacent mesentemic collateral vessels and the gallbladder bed. In one patient, collateral vessels could be seen extending from the gallbladder to the anterior abdominal wall (Fig 3a, 3b). In all four patients in whom flow-related MR sequences were used, flowrebated enhancement could be identifled within the gallbladder wall (Fig
Images
of a 12-year-old
US
scan
boy
through
wall (arrows). the gallbladder
Detectability
the
with
chronic
gallbladder
Note-AS = vanices adequately tion not performed. * PVT = portal vein thrombosis, 1 These patients had surgically
hepatitis
serpentine
image
and anechoic
corresponding
hemobilia. areas
within
the
to a demonstrates
Wall Vanices
(y)
1/M/59 2/M/35t 3/Ff771 4/M/7t 5/M/63 6/M/66
active
shows
(b) Color Doppler flow wall vanices (arrows).
of Gallbladder
Patient/Sex/Age
hypo-
echoic band (in one patient) in the gallbladder wall. They could actually be linked to extracholecystic vessels in one case. A random sampling of 10 patients with no known liver or biliamy disease revealed no sonographically
Figure
(a) Transverse gallbladder flow within
Diagnosis
CT
US
MR
Cirrhosis Biliary cirrhosis Pancreas carcinoma Chronic active hepatitis Cirrhosis Cirrhosis
AS AS NS NP AS AS
AS NP WS WS WS AS
NP WS AS AS AS NP
seen.
WS
varices
+ = pnesent. proved gallbladder
well
wall
seen.
NS
varices
not seen.
PVT* + + + + + ...
NP
examina-
vanices.
the
3). Signal
wall echo only two gested only
void
within
the
gallbladder
was also identified with spinTi-weighted sequences, but in two of these patients. Of the angiognaphic studies, one sugportal vein thrombosis, but a diffuse
tangle
of vessels
in the
right upper quadrant was noted and could not confirm gallbladder vances. Although the group assignments represent a subjective assessment, we believed that color Doppler flow was superior to MR and CT in depicting gallbladder wall vamices. 180
#{149} Radiology
DISCUSSION Gallbladder erately unusual
wall vanices manifestation
tal hypertension (12%)
Associated
portal
was present In a previous
gallbladder also
noted
was
not
vanices, that
seen.
First, resent
The
the
presence
in five case
in this
vein
throm-
of six report
et ab (18) portal vein presence and boca-
as well as the disallow us to sugexplanations for
of gallbladder
vein
paof
Rails
and most likely, a portosystemic
cystic
in six
cases
a patent
tion of thrombus, tribution of varices, gest three possible
the
a modof por-
occurred
of 50 consecutive
series. bosis tients.
and
are
branch
vamices.
portal
system to systemic anterior abdominab wall collatemals. Second, they may act as a bypass around a focally
thrombosed the the
portal cystic
extrahepatic vein. vein
Retrograde to gallbladder
segment
of
flow from vanices
to flow across the galldirectly into the hepatic and ultimately into the vein. Third, gallbladder merely represent dilated
veins due to simple back-pressure within the portal venous system. Regardless of their cause, gallbladder wall vanices have received only limited attention in the radiography literature (18-20). This in part stems from their poor visualization on angiographic studies. Although only
two
they may repshunt linking
of the
may give rise bladder bed parenchyma right portal varices may
of our
patients
underwent
angi-
ography, they exemplify the difficulty in confirming the presence of these small vanices. The angiographic detection of portal venous collatemals relies on their indirect opacification during the portal phase of a mesentemic artery injection. Because of their small caliber and the large number of collateral vessels in the porta hepatis and hepatoduodenal ligament, gallbladder varices may be obscured and
April
1991
b. Figure
2.
(a) Contrast-enhanced
axial
C-
CT image
of a 77-year-old woman with adenocarcinoma of the pancreas shows irregular thickening and enhancement in the gallbladder wall (long straight arrows), a markedly dilated common bile duct (short straight arrow) with adjacent collateral vessels (arrowheads), and a mass in the head of the pancreas (curved arrow). (b) Longitudinal sonognam through the gallbladder reveals only thickening of the gallbladder wall (arrows). (c) Color Doppler flow image in the same patient photographed in gray scale shows venous flow within the gallbladder wall, confirming that the gallbladder wall thickening is secondary to vanices (arrows) (cf b).
I .‘,
4t
.
S
O
I
,t
%
b.
c.
Figure 3. (a) Contrast-enhanced axial CT image from a 35-year-old man with primary biliary cirrhosis shows serpentine enhancement within the gallbladder wall and penicholecystic fat (arrows). (b) Image obtained 2 cm cephalad shows irregular enhancement within the gallbladder wall (arrowheads) and adjacent collateral vessels coursing toward the anterior abdominal wall (arrows). (c) Coronal gradientecho MR image of the same patient obtained with 50/12 and 45#{176} flip angle reveals flow-related enhancement in gallbladder wall varices rowheads), cavernous transformation of the portal vein (straight arrows), and vanices within the gastric wall (wavy arrow).
escape
detection. MR imaging all proved capable of depicting gallbladder varices. In this small series,
Us,
angiographic
CT, and
Doppler
US allowed
the
most
by acute
cholecystitis,
hepati-
tis, or ascites (21). The anechoic tend to be more serpentine and occasionally cystic vessels.
be
linked Because
to extracholeincreased ante-
nab flow may be observed bladder wall in patients cholecystitis, Doppler
Volume
flow
179
areas
can
in the gallwith acute
both duplex and color imaging are essential
Number
#{149}
I
confirming
the
diagnosis.
CT showed
appears
The
normal gallbladder wall should demonstrate detectable venous as would be found in vanices.
confi-
dent diagnosis of vanices. A thickened gallbladder wall containing anechoic areas may be seen with vances. This appearance is similar to that
caused
for
enhancement
not flow,
of the
thickened gallbladder wall in four patients. The enhancement pattern differed from that seen in the normal gallbladder or in acute or chronic cholecystitis. Vamices appeared as nodular enhancement within the gallbladder wall due to vessels cross-
ing
perpendicular
Numerous
were also seen bladder in the Normal
within
to the
enhancing
plane.
gallbladder
may
wall,
be
smoother
evidence
band
within
present, lecystitis.
fuse
the
vessels
due
and
the
gallbladder
enhancing
wall
stranding
The
less enhancement and is more random around (especially proved
was
in choto con-
pericholecystic
mesentemic
to inflammation.
entation imaging imaging)
uniform.
as might be expected It is important not with
to show varices
more
of a low-attenuation
latter
tends
than do in its on-
the
gallbladder. MR flow-enhanced capable of depicting
gallfat.
gallbladder vanices but did not add any information to the CT or sonographic findings despite the avail-
seen
ability
vessels
adjacent to the pericholecystic
enhancement
the
scan
small
No
(an-
but
it
The
of coronal
imaging.
radiologist
plays
a major Radiology
role #{149} 181
in distinguishing gallbladder wall vances from other causes of gallbladden wall thickening. In patients with right-upper-quadrant pain and portal hypertension, anechoic or hypoechoic areas in the gallbladder wall at US cannot be assumed to be due to inflammation. In patients with portal hypertension undergoing biliary bypass on cholecystectomy, the surgeon benefits from the preoperative knowledge that significant blood loss may occur from gallbladder varices (22). Similarly, if liver transplantation is considered, the knowledge that gallbladder vanices are present will reduce handling of the gallbladden prior to achieving adequate vascuban control. In conclusion, gallbladder wall varices can mimic a variety of diseases. Their presence may imply portal vein thrombosis. Although this represents only a small series, colon Doppler flow imaging appears to be the most sensitive and specific for the detection of these varices and should be used in patients with known or suspected portal hypertension undergoing examinations of the liven and biliary tree, particularly as a part of preoperative planning. U
References
12.
LE. Case report: acute portal vein thrombosis. Clin Radiol 1987; 38:645.
1.
Albertyn
2.
Lee WB, thrombosis Australas
3.
4.
5.
6.
7.
8.
CT demonstration of venous system. Radiol 1988; 32:360-362. Kim YC, Handler SJ, Conroy WV. Multiplanar demonstration of spontaneous portal-systemic shunts by ultrasound and computed tomography. J Comput Assist Tomogn 1980; 4:10-18. Dach JL, Hill MC, Pelaez JC, LePage JR, Russell E. Sonography of hypertensive portal venous system: correlation with an-
10.
11.
KP.
of the portal
tenial
portography.
Kane
RA, Katz
AJR
1981;
vessels
SG.
to
respiration.
sound
of arterial
scanning.
portography
Ciin Radiol
14.
hausen
15.
16.
17.
18.
and
FE.
Portal
hypertension
evaluated
by MR imaging. Radiology 1985; 157:703706. Alpern MB, Rubin JM, Williams DM, Capek P. Porta hepatis: duplex Doppler US with angiographic correlation. Radiology 1987; 162:53-56. Patniquin H, LaFortune M, Burns PN, Dauzat M. Duplex Doppler examination in portal hypertension: technique and anatomy. AJR 1987; 149:71-76. Miller yE, Benland LL. Pulsed Doppler duplex sonography and CT of portal vein thrombosis. AJR 1985; 145:73-76. RaIls PW, Mayekawa DS, Lee KP, et al. Gallbladder wall varices: diagnosis with color flow Doppler sonography. JCU 1988; 16:595-598.
19.
Saigh JC.
20.
Radiology
1982; 142:176-172. DiCandio G, Campatelli A, Mosca F, et al. Ultrasound detection of unusual spontaneous portosystemic shunts associated with uncomplicated portal hypertension. J Ultrasound Med 1985; 4:297-305. Fnider B, Mann AM, Goldberg A. Ultrasonographic diagnosis of portal vein cayernous transformation in children. J Ultrasound Med 1989; 8:445-449. Raby N, Karani J, Powell-Jackson R, et al. Assessment of portal vein patency: companison
13.
137:511-517.
The spectrum of sonographic findings in portal hypertension: a subject review and new observations. Radiology 1982; 142:453-458. Subramanyam BR, Balthazar EJ, Madamba MR, et al. Sonognaphy of portosystemic venous collaterals in portal hypertension. Radiology 1983; 146:161-166. Juttner HU, Jenney JM, Rails PW, Goldstein LI, Reynolds TB. Ultrasound demonstnation of portosystemic collaterals in cirrhosis and portal hypertension. Radiology 1982; 142:459-463. Bolondi L, Gandolfi L, Arienti V. et al. Ultrasonography in the diagnosis of pontal hypertension: diminished response of
portal 9.
Wong
Zirinsky K, Markisz JA, Rubenstein WA, et al. MR imaging of portal vein thrombosis: correlation with CT and sonography. AJR 1988; 150:283-288. Levy HM, Newhouse JH. MR imaging of portal vein thrombosis. AJR 1988; 151:283-286. Williams DM, Cho KJ, Aisen AM, Eck-
21.
22.
J, Williams Vanices:
5, Cawley
a cause
of focal
K, Anderson gallbladder
wall thickening. J Ultrasound Med 1985; 4:371-373. Salam AA, Goldman M, Smith D, Hill HL. Gastric, intestinal, and gallbladder vances: hemodynamic and therapeutic considenations. South Med J 1979; 72:402-408. Cohan RH, Mahony BS, Bowie JD, et al. Striated intramural gallbladder lucencies on US studies: predictors of acute cholecystitis. Radiology 1987; 164:31-35. Hymes JL, Jaicken BN, Schein CJ. Vanices of the common bile duct as a surgical hazand. Am Sung 1977; 43:686-688.
ultra-
1988; 39:381-
386.
182
#{149} Radiology
April
1991