AJR 2002; 178:362-363
© American Roentgen Ray Society
Transjugular Intrahepatic Portosystemic Shunt and Transjugular Embolization of Bleeding Rectal Varices in Portal Hypertension
Nico Hidajat1,
Hanno Stobbe1,
Norbert Hosten1,
Ralf-Juergen Schroeder1,
Michaela Fauth2,
Thomas Vogl3 and
Roland Felix1
1
Department of Radiology, Charité, Virchow
Clinic of the Humboldt University, Augustenburger Platz 1,13353 Berlin,
Germany.
2
Department of Internal Medicine, Hospital in Zehlendorf, Gimpelsteis 9, 14165
Berlin, Germany.
3
Department of Radiology, Institute for Diagnostic and Interventional
Radiology, Johann Wolfgang Goethe University, Theodor-Stern-Kai 7, 60596
Frankfurt, Germany.
Received December 11, 2000;
accepted after revision May 29, 2001.
Address correspondence to N. Hidajat.
Introduction
In portal hypertension, varices can arise in the distal esophagus and in
the gastric fundus or cardia but can also develop in the rectal plexus. Today,
the transjugular intrahepatic portosystemic shunt (TIPS) is the most frequent
remedy for recurrent variceal bleeding in patients with portal hypertension in
whom endoscopic treatment is insufficient
[1]. The combination of TIPS
and embolization has been described as very satisfactory in the prevention of
recurrent bleeding from esophagogastric varices
[2]. In this article, we
describe our experience with TIPS and transjugular embolization of rectal
varices in two patients.
Case Reports
An 86-year-old woman with cirrhosis of the liver due to chronic hepatitis C
presented at the clinic with rectal varices. In 1998, she underwent a
hemorrhoidectomy. In January 1999, the patient experienced two episodes of
anorectal bleeding; second-degree hemorrhoids and third-degree esophageal
varices were found on endoscopy. In May 1999, bleeding from third- to
fourth-degree esophageal varices occurred. The bleeding was stopped with
ethoxysclerol. In June and July 1999, the patient again experienced anorectal
bleeding. At colonoscopy, massive rectal varices were diagnosed as the source
of the bleeding. After a transfusion of 6 U of packed RBC, the patient's
hemoglobin level increased to 9.2 g/dL. No clinical sign of encephalopathy was
found.
Our second patient was a 48-year-old man with cirrhosis of the liver caused
by cardiomyopathy who presented with rectal varices. In August 1996, the
patient noticed tarry, diarrhealike stools, and in January 1997, he
experienced severe bleeding from the rectal varices. A perianal ulcer was
found on endoscopy and was assumed to have arisen because of the endoscopic
therapy for the rectal varices.
In both patients, endoscopic banding therapy could have stopped the
bleeding from the rectal varices, but this treatment was considered
insufficient to prevent further bleeding episodes. Therefore, TIPS creation
was seen to be the therapy of choice. The technique of transjugular puncture
of the portal vein was that which is usually used in the placement of a TIPS.
A 4-French, 80-cm—long catheter (Cobra; Terumo, Leuven, Belgium) was
introduced over a guidewire into the inferior mesenteric vein and the superior
hemorrhoidal vein to establish the location of hemorrhoids precisely and to
verify that the hemorrhoids communicated with the portal system. For the TIPS
placement in the woman, a single stent (Memotherm; Angiomed, Karlsruhe,
Germany) with a diameter of 10 mm and a length of 5 cm was implanted. A
1.5-m—long catheter (MicroFerret-18; Cook, Bjaerverskov,
Denmark)—with an outer diameter of 3.0 French in the proximal part and
2.4 French in the distal part—was introduced into the Cobra catheter to
add additional length to the catheter system so that the varices could be
reached. In the TIPS procedure in the man, two stents (Memotherm; Angiomed)
with diameters of 12 mm and lengths of 3 cm were implanted. In this patient,
the Cobra catheter could be introduced directly into the varices.
We performed the coil embolization procedure until variceal opacification
became markedly reduced or totally absent when a contrast agent was injected
into the superior hemorrhoidal vein. For embolization of the rectal varices,
five coils (Tornado; Cook) with diameters of 5 mm were implanted in the woman,
and 12 coils with diameters of 5 and 8 mm were implanted into the proximal
part of the varices in the man. In the woman, the esophageal varices were also
embolized. Two coils with diameters of 5 mm were used.
The superior hemorrhoidal vein and rectal varices could be seen indirectly
during the venous phase of the inferior mesenteric arteriography. The TIPS
were placed successfully, resulting in a reduction of the portosystemic
pressure gradients to 14 and 10 mm Hg in the woman and man, respectively.
After the TIPS placement, perfusion of the rectal varices in both patients,
although decreased, was still substantial and was visualized with the direct
injection of contrast agent into the inferior mesenteric vein
(Fig. 1A). Therefore, we
decided to embolize varices (Fig.
1B). After embolization, no varices could be seen when contrast
material was injected via the superior hemorrhoidal vein
(Fig. 1C). No complications
related to the procedure occurred. During the 6-month follow-up period,
neither patient had recurrent bleeding from the rectal varices.
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —86-year-old woman with cirrhosis of the liver attributable to
chronic hepatitis C. Superior hemorrhoidal venogram obtained through
selectively positioned catheter (arrowhead) shows rectal varices
(arrow).
|
|
View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —86-year-old woman with cirrhosis of the liver attributable to
chronic hepatitis C. Venogram shows microcatheter (arrowhead)
introduced into proximal part of rectal varices so coils may be placed.
|
|
View larger version (152K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —86-year-old woman with cirrhosis of the liver attributable to
chronic hepatitis C. Superior hemorrhoidal venogram was obtained after
placement of coils (arrow). Note that no rectal varices are
visualized.
|
|
Discussion
Sigmoidoscopic data show that between 44% and 90% of patients with portal
hypertension have anorectal varices
[3,
4]. About 10% of the patients
with anorectal varices have clinically consequential bleeding attributable to
the varices [3].
In 1993, Katz et al. [5]
described the first patient who was successfully treated with TIPS placement
for recurrent bleeding of anorectal varices despite having had previous rubber
band ligation. The patient had no recurrent bleeding over the ensuing 6
months. Fantin et al. [6] gave
a similar report on a patient in 1996. In that patient, no further rectal
bleeding episodes were noted during the 12 months after the TIPS placement. In
1997, Godil and McCracken [7]
described another patient with rectal variceal bleeding. Although no recurrent
bleeding occurred, rapid decompensation of liver function and encephalopathy
developed, and the patient died approximately 4 weeks after the TIPS
procedure. The largest series of patients (n = 7) with bleeding
anorectal varices (n = 7) and parastomal varices (n = 5)
treated by TIPS placement was reported by Shibata et al.
[8] in 1999. Four of these
patients had recurrent bleeding within 11 months after the TIPS placement. In
each of these patients, duplex sonography revealed occlusion of the TIPS. None
of the authors made any mention about embolization.
According to a study by Rössle et al.
[1], substantial variceal
perfusion was observed in several patients despite a reduction of the pressure
gradient to less than 12 mm Hg. Therefore, the risk of recurrent bleeding
might be lower when the pressure gradient is reduced to a lower value, as was
the case in the patient reported by Godil and McCracken
[7] in whom the gradient was 4
mm Hg.
Perhaps the probability of encephalopathy and hepatic failure increases
when the pressure gradient drops in response to the reduced hepatic perfusion
caused by the diversion of portal venous blood. In our opinion, this increased
probability is an important reason not to reduce the pressure gradient too
much. The risk of recurrent bleeding has already been somewhat reduced by the
TIPS placement. The risk can be decreased further by embolization of the
varices [2]. Another reason for
embolizing the varices is that the pressure gradient can increase
spontaneously; shunt stenosis can occur from proliferation of pseudointimal
granulation tissue or shunt occlusion, leading to redevelopment of varices.
After embolization, however, the communication of the portal vein and the
rectal veins remains partially interrupted, so that the increase of the
pressure in the portal vein is not directly transmitted into the rectal
plexus. Nevertheless, a revision of the stenotic or occluded shunt should be
performed to resolve other complications such as ascites.
Embolization of esophagogastric varices has been described by several
authors [1,
2]. Substantial variceal
perfusion despite reduction of the portosystemic pressure gradient to less
than 12 mm Hg can be seen as an indication for embolization
[1]. To our knowledge, no case
of rectal varices treated with embolization has been reported by other
authors. Our results show that bleeding rectal varices in portal hypertension
can be treated with TIPS placement and additional transjugular embolization.
Transjugular embolization should be considered in those patients in whom TIPS
placement is performed to treat such varices. We treated the rectal varices in
the same way we would treat esophagogastric varices. However, because of our
limited number of patients, more research is needed.
References
-
Rössle M, Haag K, Ochs A, et al. The
transjugular intrahepatic portosystemic stent-shunt procedure for variceal
bleeding. N Engl J Med
1994;330:165
-171[Abstract/Free Full Text]
-
Hidajat N, Vogl T, Stobbe H, et al. Transjugular intrahepatic
portosystemic shunt: experiences at a liver transplantation center.
Acta Radiol
2000;41:474
-478[Medline]
-
Hosking SW, Smart HL, Johnson AG, Triger DR. Anorectal varices,
hemorrhoids, and portal hypertension. Lancet
1989;1:349
-352[Medline]
-
Chawla Y, Dilawari DB. Anorectal varices: their frequency in
cirrhotic and non-cirrhotic portal hypertension. Gut
1991;32:309
-311[Abstract/Free Full Text]
-
Katz JA, Rubin RA, Cope C, Holland G, Brass CA. Recurrent bleeding
from anorectal varices: successful treatment with a transjugular intrahepatic
portosystemic shunt. Am J Gastroenterol
1993;88:1104
-1107[Medline]
-
Fantin AC, Zala G, Risti B, Debatin JF,
Schöpke W, Meyenberger C. Bleeding anorectal
varices: successful treatment with transjugular intrahepatic portosystemic
shunting (TIPS). Gut
1996;38:932
-935[Abstract/Free Full Text]
-
Godil A, McCracken JD. Rectal variceal bleeding treated by
transjugular intrahepatic portosystemic shunt. J Clin
Gastrenterol 1997;25:460
-462[Medline]
-
Shibata D, Brophy DP, Gordon FD, Anastopoulos HT, Sentovich SM,
Bleday R. Transjugular intrahepatic portosystemic shunt for treatment of
bleeding ectopic varices with portal hypertension. Dis Colon
Rectum 1999;42:1581
-1585[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Case Reports
