AJR 2000; 174:132-134
© American Roentgen Ray Society
Intraperitoneal Portal Vein Bleeding During Transjugular Intrahepatic Portosystemic Shunt
Treatment with Stent-Graft Placement
Elias N. Brountzos1,
Efthymia Alexopoulou1,
Ioannis Koskinas2,
Loukas Thanos1,
Matthildi A. Papathanasiou1 and
Dimitrios A. Kelekis1
1
Second Department of Radiology, Medical School, Athens University, Eugenidion
Hospital, 20 Papadiamantopoulou St., Athens 11528 Greece.
2
Academic Department of Medicine, Hippokration Hospital, 114 Vas. Sofias St.
Athens 11527, Greece.
Received April 5, 1999;
accepted after revision June 22, 1999.
Address correspondence to E. N. Brountzos.
Introduction
The most difficult step in the creation of a transjugular intrahepatic
portosystemic shunt (TIPS) is the transhepatic puncture of the portal vein
because it is essentially a "blind" puncture. In the rare event of
an extrahepatic portal vein puncture, dilatation can cause free portal vein
laceration and exsanguination
[1]. A case of extrahepatic
portal vein puncture with life-threatening intraabdominal hemorrhage was
treated with a stent-graft.
Case Report
A 68-year-old woman was admitted because of hematemesis and hepatic
encephalopathy. She had a 4-year history of chronic hepatitis C infection with
Child-Pugh classification B cirrhosis and portal hypertension, and had been
hospitalized three times for gastrointestinal bleeding. Endoscopy showed
severe hypertensive gastropathy with active bleeding in the antrum. Bleeding,
requiring transfusion of 10 units of blood, persisted during hospitalization.
The patient was referred for a TIPS procedure.
Preinterventional abdominal CT, liver sonography, and color Doppler
sonography showed that the main right and left portal veins were patent, as
were the right and left hepatic veins. Antibiotics were started 24 hr before
the procedure (IV cefuroxime, 1 g/24 hr). The procedure was performed with the
patient under local anesthesia (lidocaine hydrochloride, 1%) and IV sedation
(midazolam, 2 mg; fentanyl citrate, 50 µg).
The right internal jugular vein was punctured and a 10-French, 41-cm
Check-Flo II introducer sheath (William Cook Europe, Bjaeverskov, Denmark) was
advanced into the superior vena cava. Using a 5-French selective catheter
(Torcon NB; William Cook Europe) we catheterized the right hepatic vein. A
wedged hepatic venogram using a hand injection of 20 ml of undiluted contrast
material failed to depict the portal vein. Gaseous carbon dioxide was not
available at the time. With the preprocedural imaging as a guide, we attempted
puncture of the portal vein. We performed a puncture in the proximal 4 cm of
the hepatic vein with a Rosch-Uchida Transjugular Access set (William Cook
Europe). During the third attempt, the portal system was entered at the point
assumed to be the right portal vein just proximal to bifurcation. Pressure
measurements revealed a portal-to-systemic gradient of 3 cm of water.
Portography through a 5-French pigtail catheter (Torcon NB) showed patent
portal and splenic veins without evidence of contrast extravasation
(Fig. 1A).
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —68-year-old woman with Child-Pugh B classification cirrhosis and
portal hypertension was referred for TIPS placement because of upper
gastrointestinal bleeding and ascites.
A, Transhepatic portal venogram shows portal and hepatic veins are
patent. Note lack of evidence of contrast material extravasation. Portal
system entrance point was assumed to be right portal vein just proximal to
bifurcation.
|
|
The parenchymal tract was subsequently dilated using an 8 x 40 mm
balloon catheter (Accent; William Cook Europe). A second portogram was then
obtained that showed brisk extravasation of contrast material from the portal
vein into the hepatic hilum. The portal vein puncture was located
extrahepatically (Fig. 1B). At
this point the patient developed signs of hypovolemic shock with a blood
pressure of 70 over 40 mm Hg and a pulse rate of 150 beats per minute. The
hypovolemic shock was treated with rapid IV infusion of colloids.
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —68-year-old woman with Child-Pugh B classification cirrhosis and
portal hypertension was referred for TIPS placement because of upper
gastrointestinal bleeding and ascites.
B, Transhepatic portal venogram after dilation of parenchymal tract
with an 8x40 mm balloon shows brisk extravasation of contrast material
(arrows) from portal vein in hepatic hilum.
|
|
We introduced a 10 x 40 mm Passager stent-graft (Boston Scientific,
Watertown, MA) consisting of a self-expanding nitinol frame covered with a
0.1-mm-thick polyester fabric with a low porosity of 140 ml of water per
minute per square centimeter. We positioned it bridging the portal puncture
with its distal end approximately 2 cm into the main portal vein. The proximal
end of the stent-graft was located 2 cm above the portal bifurcation in the
parenchymal tract.
The patient immediately stabilized with marked reduction in fluid
requirements. Subsequently a 10 x 90 mm Wallstent (Schneider-Europe,
Bulach, Schwitzerland) was placed inside the stent-graft to expand the entire
parenchymal tract. The proximal end of the Wallstent was placed in the right
hepatic vein. The stent and stent-graft were then dilated with a 10 x 40
mm balloon. A portogram showed that the extravasation had stopped and there
was minimal filling of the intrahepatic portal veins
(Fig. 1C). Pressure
measurements revealed a gradient of 10 cm of water; therefore, the procedure
was terminated.
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —68-year-old woman with Child-Pugh B classification cirrhosis and
portal hypertension was referred for TIPS placement because of upper
gastrointestinal bleeding and ascites.
C, Transhepatic portogram shows 10x40 mm stent-graft
(arrows) being placed to bridge portal puncture while 10 x 90
mm stent (arrowheads) is placed inside graft to expand entire tract.
Note that extravasation of contrast material has stopped.
|
|
At the end of the procedure, the patient was hemodynamically stable.
Because her hematocrit was 15%, she was given a transfusion of four units of
packed RBCs. The transfusion resulted in a hematocrit of 25%, which remained
stable thereafter. Her melena stopped and a second endoscopy showed no
evidence of portal gastropathy. The patient remained free of bleeding 7 months
after the procedure, when a second color Doppler sonogram and venogram showed
a patent TIPS tract and intrahepatic portal veins
(Fig. 1D).
View larger version (159K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1. —68-year-old woman with Child-Pugh B classification cirrhosis and
portal hypertension was referred for TIPS placement because of upper
gastrointestinal bleeding and ascites.
D, Transjugular portogram 7 months after procedure shows patent TIPS
tract and intrahepatic portal veins.
|
|
Discussion
Extrahepatic portal vein laceration during TIPS has been previously
reported in the literature [1,
2,
3,
4].
When a TIPS procedure is performed, the portal vein puncture should be
located intrahepatically at least 1-2 cm above the portal bifurcation
[1]. Anatomic studies have
shown that almost 50% of the portal vein bifurcations in patients undergoing
TIPS procedures are extrahepatic
[3]. Therefore, when in doubt,
radiologists should abandon this access and attempt a new puncture in an
intrahepatic location.
Alternatively, Freedman and Sanyal
[1] treated a free portal vein
rupture successfully by placing a bare Wallstent. Similarly Davis and Haskal
[4] described two cases of
extrahepatic portal vein puncture successfully treated with insertion of
uncovered stents. Krajina et al.
[3] reported three patients
with this complication. In one patient, the insertion of a bare Wallstent
failed to seal the laceration and the patient died. In the two remaining
patients, bleeding was successfully treated using stainless steel stents
covered with polyester fabric. We effectively treated this serious
complication with stent-graft placement. Ideally, we could have used a longer
stent-graft, but the 40-mm device was the only one available. Because the
stent-graft did not cover the entire parenchymal tract, we used a Wallstent to
complete the TIPS procedure.
We were concerned about the long-term patency of the TIPS in our patient.
Krajina et al. [3] reported
that graft thrombosis developed 5 months after the procedure in one patient,
whereas in the remaining patient stenosis occurred at the nongrafted portion
of the tract. Our patient's venogram 7 months after the procedure showed only
minimal intimal thickening without significant stenosis. In animal studies,
the use of polytetrafluoroethylene-covered stent-grafts resulted in increased
TIPS patency [5,
6]. In a different animal
study, polyethylene terephthalate (Dacron; duPont, Wilmington, DE) covered
with polyurethane was used as a graft material
[7]. The use of this material
did not improve TIPS patency. In humans, polytetrafluoroethylene-covered
stent-grafts have been used to revise a TIPS stenosis and have been found to
increase long-term patency
[8].
Although we used a polyester-covered stent-graft, which is associated with
early stenosis both in animal and human studies
[3,
7], the TIPS remained patent 7
months after the procedure. This result may be attributed to the use of a
relatively short stent-graft that only partially covered the TIPS tract. The
short stent-graft may also have accounted for the preservation of the patency
of the intrahepatic portal veins. One potential complication of the use of
stent-grafts in TIPS placement is the occlusion of the intrahepatic portal
veins or the hepatic veins [8].
The rate of fatal complications is inversely related to the number of TIPS
procedures performed [1]. This
case occurred early in our experience with TIPS. Nevertheless, our experience
with TIPS. Nevertheless, our experience with stent-grafts in other settings
and the availability of the stent-graft allowed successful treatment of this
potentially fatal complication. We continue to use bare metal stents for
creation of a TIPS, with careful localization of the portal bifurcation, but
keep a stent-graft available.
Acknowledgments
We thank John A. Kaufmann, associate professor of radiology, Harvard
Medical School, for his invaluable help in the preparation of the
manuscript.
References
-
Freedman A, Sanyal A. Complications of transjugular intrahepatic
portosystemic shunt. Semin Intervent Radiol 1994;
11:161-177
-
Rossle M. Puncture of the portal bifurcation: a fatal complication
of TIPS (letter). Radiographics 1993;
13:1184
-
Krajina A, Hulek P, Ferko A, Nozicka J. Extrahepatic portal venous
laceration in TIPS treated with stent graft placement.
Hepatogastroenterology 1997;
44:667-670[Medline]
-
Davis AG, Haskal ZJ. Extrahepatic portal vein puncture and
intra-abdominal hemorrhage during transjugular intrahepatic portosystemic
shunt creation. J Vasc Interv Radiol 1996;
7:863-866[Medline]
-
Nishimine K, Saxon RR, Kichikawa K, et al. Improved transjugular
intrahepatic portosystemic shunt patency with PTFE-covered stent-grafts:
experimental results in swine. Radiology 1995;
196:341-347[Abstract/Free Full Text]
-
Haskal ZJ, Davis A, McAllister A, Furth EE. PTFE-encapsulated
endovascular stent-graft for transjugular portosystemic shunts: experimental
evaluation. Radiology 1997;
205:682-688[Abstract/Free Full Text]
-
Bloch R, Pavcnik D, Uchida BT, et al. Polyurethane-coated
Dacron-covered stent-grafts for TIPS: results in swine. Cardiovasc
Intevent Radiol 1998;
21:497-500[Medline]
-
DiSalle RS, Dolmatch BL. Treatment of TIPS stenosis with ePTFE
graft-covered stents. Cardiovasc Intervent Radiol
1998;
21:172-175[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Case Report
