You are on page 1of 2

Gastrointest Radiol 17:141-144 (1992)

Gastrointestinal
Radiology
9 Springer-Verlag New York Inc. 1992

Portal Vein Reflectors: Duplex Sonographic Appearance


Janis Gissei Letourneau, John E. Carlson, Deborah G. Longley, Joseph W. Yedlicka, Jr.,
and Wilfrido R. Castafieda-Zt~fiiga
Department of Radiology. University of Minnesota. Minneapolis, Minnesota. USA

Abstract. Duplex sonography has established utility shunts; additionally, vascular complications of he-
in the noninvasive evaluation of the portal venous patic surgery, particularly subtotal hepatectomy and
system. Recently, the duplex sonographic features liver transplantation, can be defined with this nonin-
of suspected portal venous air have been described. vasive modality [I-13]. Real-time gray-scale and du-
We report on an experimental study in dogs under- plex sonographic features of air within the portal
taken to establish if the same sonographic features vein in native and transplanted livers have been de-
could be reproduced in a laboratory setting and to scribed [14-17].
determine if small fragments of clot emboli could Sonographic features of portal venous air include
produce similar gray-scale and Doppler findings. In- intermittently visible high-amplitude echoes within
jections of microbubbles of air and clot fragments the portal vein that move in the direction of blood
into the portal venous system were monitored using flow. Poorly defined echogenic foci can also be seen
duplex ultrasound. The gray-scale and Doppler fea- within the hepatic parenchyma in nondependent re-
tures of flow in the main portal vein were indistin- gions. These small echogenic foci, thought to repre-
guishable in both microbubble and clot injections. sent air bubbles, produce characteristic Doppler
Superimposition of high-amplitude spikes on the spectra with spike-like "blips" superimposed on the
normal portal venous waveform was seen in 19 of 24 normal, even and continuous signal from red blood
(79%) microbubble and t3 of 23 (56%) clot fragment cells flowing within the portal vein.
injections. We conclude that the appearance of rap- We undertook an experimental study in dogs to
idly moving, bright intraluminal echoes coupled with verify these duplex sonographic features of portal
spike-like aberrations of the portal venous waveform venous air and to determine if microembolic autolo-
can be associated with portal venous air bubbles or gous blood clot fragments could produce similar du-
small blood clot emboli. plex findings.
Key words: Portal vein, g a s l P o r t a l vein, US stud-
ies--Portal vein, abnormalities.
Materials and Methods

Eight mongrel dogs (25-35 kg) were lightly anesthetized with in-
travenous sodium thiamylal. The dogs were then intubated, venti-
Duplex sonography is currently a widely accepted lated, and monitored throughout the experiment. Through a right
technology for evaluation of the portal venous sys- subcostal incision, small intestine and superior mesenteric venous
tem. Diagnostic applications exist for determining branches were exposed. A branch of the superior mesenteric vein
vascular patency, direction of flow, and presence of within the small bowel mesentery was cannulated with an 18-
gauge intravenous catheter (Fig. I).
portosystemic collaterals. Clinical settings in which Preliminary examination of the main portal vein using duplex
this technology is useful include portal venous hy- Doppler ultrasound (5.0-MHz sector transducer, Acuson, 128,
pertension, malignancy, and surgical portosystemic Acuson Computed Sonography, Mountain View, CA, USA) was
performed to verify patency. Microbubbles of air, with an average
diameter of 75/zm, were produced by to-and-fro syringe agitation
Address ~(fprint requests to: Janis Gissel Letourneau, M.D.. De- of 50% dextrose [18]. The suspensions of microbubbles were then
partment of Radiology, University of Minnesota Hospital, 420 injected at 10 ml/min through the intravenous catheter. All injec-
Delaware Street S.E., Box 292 UMHC. Minneapolis, MN 55455. tions were directly monitored by duplex sonography of the main
USA portal vein or its major branches using the liver as an acoustic
142 J.G. Letourneau et al.: Portal Vein Reflectors

flowing, bright intraluminal echoes within the main


portal vein on real-time scanning in every instance;
these bright intraluminal echoes were innumerable,
filling the lumen of the portal vein. The gray-scale
appearances of these different intraluminal echoes
were indistinguishable (Figs. 2B and 3).
Simultaneous Doppler examination of the main
portal vein during these injections showed superim-
position of high-amplitude, narrow spikes, or
"blips," on the expected, normal portal venous
waveform with 19 of 24 (79%) air microbubble injec-
tions and 13 of 23 (56%) clot fragment injections.
Spectral waveforms of air microbubble and clot frag-
ment injections were also indistinguishable (Figs. 2
and 3).
Control injections occasionally produced spo-
radic luminal echoes in the portal vein lumen; this
was seen in five of the 19 control injections and, in
all instances, was strikingly less prominent than the
appearance of intraluminal echogenicity seen during
experimental injections. In the first control animal a
small amount of air bubbles was lbund in the injec-
tion system and thought to have been completely
cleared from the line: it is possible that minimal air
Fig. 1. Schematic diagram showing surgical exposure and cannu- remained in the line for the remainder of injections
lation of a branch of the superior mesenteric vein. Duplex exami- with this animal as three of the injections (two sa-
nation of the main portal vein was performed by scanning with the
line, one dextrose) resulted in occasional moving in-
transducer directly on the surface of the liver.
traluminal echoes. In only one instance were any
aberrations noted in the Doppler spectrum, an in-
stance of injection of normal saline; this was seen on
window, scanning directly on the hepatic surface. Because of
technical constraints on imaging of the superiorly and trans-
the first injection marred by air bubbles that were
versely oriented canine liver, it was not possible to image, or apparently drawn into the syringe inadvertently.
interrogate with Doppler, peripheral areas of the liver. Three
hand injections of microbubbles were performed on each dog with Discussion
complete clearance of gray-scale and Doppler changes between
injections (Fig. 2A). Duplex examinations were recorded with
video and film archiving. Portal venous air is an uncommon finding that is
Autologous microclots were produced by coagulating blood in seen in association with bowel infarction; inflamma-
the presence of bovine thrombin in two 6-ml syringes and frag- tory disease of the bowel, such as ulcerative colitis,
mented by suspending the clot in 50 ml of normal saline and appendicitis, cholangitis, and necrotizing enterocoli-
forcing the clot through a syringe tip. The average diameter of the
microclots, determined by the use of an hemocytometer under tis; intraabdominal abscess; small bowel obstruc-
microscopic observations, was 100 ~m. Three hand injections at tion; and gastric ulcer: and, more recently, after
10 ml/min of autologous microclots were then performed in each liver transplantation [14-17, 19-22]. In a recent re-
dog, again allowing for gray-scale and Doppler clearance of any port, seven (18%) of 39 liver transplant recipients
abnormalities between injections. In one dog, only two injections
were thought to have intraluminal portal vein air in
of microclots were done due to insufficient volume of available
microclots. the first 2 weeks after transplantation [14]. Signifi-
In total, 24 injections of air microbubbles and 23 injections of cant clinical data indicated that this duplex finding
microclots were performed. was associated with sepsis, Clostridium intestinal
Two additional control dogs were studied. The dogs were pre- colonization, postoperative ileus, and primary graft
pared identically as in the previous experiments. Solutions of
normal saline and 50% dextrose were each injected at l0 ml/min.
failure in one patient each. No associated posttrans-
A total of l0 injections of normal saline and nine injections of 50% plant complications were thought to exist in the re-
dextrose were observed in the two dogs. maining three patients. None of the seven affected
patients had clinical evidence of bowel infarction or
necrosis. The authors of this study concluded that
Results portal venous air can be commonly observed after
liver transplantation and that its presence may not
Injection of both air microbubbles and autologous portend a poor prognosis for the recipient. Similar
clot fragments caused the appearance of rapidly duplex findings have been reported in the literature

You might also like