DOI:10.2214/AJR.04.1831
AJR 2006; 186:1138-1143
© American Roentgen Ray Society
Comparative Results of Doppler Sonography After TIPS Using Covered and Bare Stents
Douglas Lake1,2,
Marcelo Guimaraes2,
Susan Ackerman1,
Christopher Hannegan2,
Claudio Schonholz2,
J. Bayne Selby2 and
Renan Uflacker2
1 Division of Ultrasound, Medical University of South Carolina, 169 Ashley Ave.,
Charleston, SC 29425.
2 Division of Interventional Radiology, Medical University of South Carolina,
Charleston, SC 29425.
Received November 29, 2004;
accepted after revision February 21, 2005.
Address correspondence to D. Lake.
Abstract
OBJECTIVE. Our purpose was to evaluate the role of sonography in the
early follow-up of patients with a covered transjugular intrahepatic
portosystemic shunt (TIPS).
CONCLUSION. Routine baseline Doppler sonography should occur
7–14 days after shunt placement unless malfunction or procedural
complications are suspected.
Keywords: color Doppler sonography • interventional radiology • liver disease • stents
Introduction
The transjugular intrahepatic portosystemic shunt (TIPS) has been a widely
accepted treatment for complications of portal hypertension
[1]. Covered stents have
recently been introduced that have shown improved patency
[2–4].
Because of its low cost and noninvasive nature, Doppler sonography is used as
a screening test in asymptomatic patients and as a noninvasive initial
evaluation in individuals with clinically suspected TIPS malfunction
[5–8].
It has been noted that the hepatic venous end of covered stents has been
difficult to evaluate with Doppler sonography
[2]. To our knowledge, this has
not been directly addressed in the literature. We performed this retrospective
study to compare initial Doppler sonography findings in TIPS performed with
expanded polytetrafluoroethylene (ePTFE)–covered stents and bare stents
to identify the differences and perhaps propose an alternative follow-up
strategy.
Materials and Methods
All 60 patients who underwent de novo TIPS procedures between December 2000
and May 2004 at our institution were considered for inclusion in our study. We
retrospectively reviewed patient records in an electronic database and PACS.
Two groups were created for analysis based on the stent type received.
Covered-stent endoprostheses (Viatorr, W. L. Gore and Associates) constituted
the first group, and noncovered endoprostheses (Wallstent, Boston Scientific)
constituted the second group. At our institution, the follow-up protocol calls
for sonography at 1–3 days, 3 months, and 6 months after TIPS placement
and at subsequent 6-month intervals. Patient compliance for follow-up
examination varied. A patient was included in the study if follow-up Doppler
sonography was performed within 4 days of TIPS.
A database was compiled that included clinical and procedural information
about all patients (i.e., age, sex, and Child-Pugh classification; indication
for the procedure; the number and diameter of stents placed; and portosystemic
pressure gradient before and after the procedure) and baseline Doppler
sonographic data (including flow velocities at the portal venous and hepatic
venous ends, number of days after TIPS, and the type of machine used).
Subsequent pathologic correlation after liver transplant and follow-up
venographic results were included if available. Patients who received no
sonographic follow-up (n = 7; 6 from the Wallstent group, one from
the Viatorr group) or whose baseline sonography occurred 4 days after TIPS
(n = 8; 5 from the Wallstent group, 3 from the Viatorr group) were
excluded from analysis. Patients who received Viatorr stents within partially
occluded Wallstents (n = 1) and patients who received Viatorr stents
and Wallstents simultaneously (n = 1) were also excluded. Thus, 43
patients were included in the study (19 with Viatorr stents and 24 with
Wallstents).
Sonographic findings at the initial examination included color Doppler
flow, waveform analysis, and measurement of velocities both at hepatic and
portal venous ends of the stent. Abnormal findings included no velocity
measurement in either the hepatic venous or portal venous end of the TIPS
(n = 9, all from the Viatorr group), gradient between proximal and
distal ends of TIPS greater than 150 cm/sec (n =4, 1 from the Viatorr
group and 3 from the Wallstent group) or distal hepatic venous shunt velocity
< 40 cm/sec (n = 1, Viatorr group). All these abnormal findings
are suspicious for malfunctioning TIPS and the lack of color flow suggests
TIPS thrombosis.
Baseline sonograms were obtained within 24 hr (n = 32), 48 hr
(n = 4), 72 hr (n = 3), or 96 hr (n = 4). All
sonograms were obtained after overnight fasting when possible. All were
performed on either GE Healthcare 900 or 700 sonography units using a
2–4 MHz phased-array transducer. All examinations were performed by or
directly supervised by an experienced radiologist.
After TIPS creation, our Doppler sonography protocol included spectral and
color Doppler examination of the stent lumen, the portal veins, and the
hepatic veins. We obtained velocities from the proximal and distal ends of the
stent. We also obtained velocities from any area of the stent from which
findings appeared abnormal on color Doppler sonography. We performed Doppler
sonography to determine flow velocity using an angle of 60° or less. Other
parameters such as pulse repetition frequency, wall filter, Doppler gain, and
transmitted frequency were adjusted to enhance sensitivity and minimize
artifact depending on body habitus and the vessel being examined. Other
specific parameters evaluated included flow velocity in the main portal vein
and the direction of flow in the left portal vein.
A subsequent paired comparison was performed between initial and
short-interim follow-up sonography among all patients with abnormal findings
on initial sonography (11/19) in the Viatorr covered-stent group. Although
patient compliance varied, a limited sample of patients (4/11) received
follow-up sonography 4–14 days after the TIPS procedure.
Statistical analysis was performed using an unpaired two-tailed student
t test to determine the differences between the Viatorr and Wallstent
groups in sex, age, Child-Pugh classification, number of stents placed,
diameter of stents placed, length of stents, portosystemic pressure gradient
before and after TIPS, pressure gradient improvement, number of days until
sonography, sonographic proximal stent lumen velocity, sonographic distal
stent lumen velocity, and categorization of normal versus abnormal sonographic
findings. Statistical analysis of the initial versus subsequent sonographic
scans was performed using a paired samples t test. SPSS, 12.0 (SPSS)
for Windows (Microsoft) was used for all statistical analysis. All statistics
are presented as mean ± SD unless noted otherwise. Differences were
considered statistically significant if p < 0.05. The study was
approved by the institutional review board of our institution.
Results
A comparison of the clinical data for the patients with Viatorr stents
versus Wallstents is shown in Table
1. No significant difference was seen between the patients with
Viatorr and those with Wallstents in age, sex, Child-Pugh classification, mean
portosystemic pressure gradient, number of stents, or total stent length. In
the Viatorr group, significant differences were noted in stent diameter
(p < 0.001) and mean portosystemic pressure gradient after TIPS
placement (p = 0.006). In the Viatorr group, 17/19 (89.5%) received
10-mm stents whereas 2/19 (10.5%) received 12-mm stents. In the Wallstent
group, 7/24 (29.2%) received 10-mm stents whereas 17/24 (70.8%) received 12-mm
stents. The mean portosystemic shunt gradient after TIPS in the Viatorr group
was 6.0 ± 3.1 versus 8.8 ± 3.3 in the Wallstent group.
A comparison of the sonographic data for the patients with Viatorr stents
versus Wallstents is shown in Table
2. No significant difference was seen between the number of days
until sonographic evaluation (1.3 days in the Viatorr group vs 1.4 days in the
Wallstent group) or portal venous shunt velocity (91.4 cm/sec vs 89.4 cm/sec,
respectively). In the Viatorr group, significant differences were noted in the
hepatic venous shunt velocity (p = 0.005) and in the overall
interpretation (p = 0.002). The average velocity of the hepatic
venous shunt was significantly lower in the Viatorr group (70.2 cm/sec vs
135.7 cm/sec in the Wallstent group) secondary to nine patients in whom no
velocity could be measured. Excluding the patients in whom final
interpretation was abnormal, the average velocity of the hepatic venous shunt
was significantly lower in the Wallstent group than in the Viatorr group
(116.0 cm/sec vs 132.4 cm/sec, respectively)
(Fig. 1). Abnormal sonographic
interpretation was seen more commonly in the Viatorr group (11/19, 57.9%) than
in the Wallstent group (3/24, 12.5%) at early sonographic evaluation.
View larger version (14K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —Scatterplot shows comparison of hepatic venous velocities between
Viatorr (W. L. Gore and Associates) group and Wallstent (Boston Scientific)
group. Data presented as mean ± SD for each group. Normal sonography
data points are shown in green and abnormal sonography data points are shown
in blue. Note nine data points at value 0 in Viatorr group.
|
|
In our final data set, the initial sonographic data of examinations
performed within 4 days of Viatorr covered-stent placement was compared with
sonographic data obtained between 4 and 14 days after TIPS using a paired
comparison. These data are shown in Table
3. Significant differences were noted between the number of days
until sonography (p = 0.02), hepatic venous shunt velocities
(p < 0.001), and overall interpretation (p < 0.001).
The number of days until sonography was the variable tested, and a significant
difference between the groups (1.3 days for Viatorr vs 10.0 days for
Wallstent, p = 0.02) was expected and is important to show. In each
of these four cases, the TIPS could be shown on gray-scale images, but no flow
could be detected on Doppler sonography. Absence of flow is abnormal, and
shunt thrombosis was suspected. However, the patients were clinically
asymptomatic, followed empirically, and follow-up sonography was normal.
Because hepatic venous velocities were detected and normalizing at follow-up,
statistically significant differences were expected among initial and
follow-up sonography, hepatic venous velocities, and overall
interpretation.
Table 4 contains follow-up
data on patients in the Viatorr covered-stent group who had abnormal initial
sonography. All available follow-up data including delayed sonography,
venography, and liver transplantation are included. The four patients included
are those from the paired comparison who had normalizing hepatic velocities at
follow-up. Of the remaining seven patients who initially had an abnormal
sonography, various methods of documenting TIPS patency were used. Two
patients had orthotopic liver transplants at 14 and 139 days, respectively,
which showed patent TIPS stents at pathologic evaluation. One patient had a
normal venogram whereas another had a four-phase liver CT, which showed TIPS
patency. Another patient had no color flow or demonstrable velocities at
either the hepatic or portal veins and died 2 days after TIPS placement. At
autopsy, this patient was noted to have complete thrombosis of the portal and
hepatic venous systems. Two patients were lost to follow-up at our medical
center.
Discussion
Post-TIPS follow-up imaging protocols vary, but sonography has been the
best screening test for TIPS malfunction because it is inexpensive,
noninvasive, and accurate. Although no consensus exists on the optimal
sonographic screening protocol
[6], most recommend a 24-hr
follow-up sonography to assess patency and establish baseline flow
measurements [9,
10]. These baseline
measurements are later used to determine temporal changes in peak shunt
velocity, which has been shown to be useful in detecting TIPS malfunction
[6,
11].
Our study was undertaken when initial sonographic examination after TIPS
with a covered stent (Viatorr) failed to show sonographic patency. In all
abnormal cases, grayscale imaging showed the shunt; however, hepatic venous
end color flow was absent, and velocity measurements could not be obtained
(Figs. 2A,
2B, and
2C). These sonographic findings
suggest a TIPS compromise such as thrombosis. Consequently, we decided to
retrospectively review a similar cohort of noncovered-stent (Wallstent)
patients who had received baseline sonography within 4 days of TIPS and
compare their results with those of similar patients who received covered
stents (Viatorr).
View larger version (126K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —45-year-old man with biopsy-proven nonalcoholic steatohepatitis.
Doppler sonogram obtained 1 day after TIPS procedure shows no flow in shunt.
Normal flow in inferior vena cava is shown.
|
|
Procedurally the Viatorr and Wallstent groups were similar except for
differences in stent diameter and postprocedure portosystemic gradient. The
difference in stent diameter was historical in nature because at our
institution we began implanting 10-mm Viatorr stents before 12-mm stents were
available. Thus, a higher percentage of 10-mm Viatorr stents were placed than
10-mm Wallstents (89.5% vs 29.2%, respectively). It has been shown in
noncovered stents that as diameter increases from 10 to 12 mm, the
postprocedure mean portosystemic gradient decreases more
[7]. This is thought to be
secondary to improved portal decompression by the larger stent. However, the
Viatorr group achieved a significantly lower postprocedure mean portosystemic
gradient despite the smaller diameter stents. To our knowledge, this has not
been described in relation to the covered versus noncovered stents and may
relate to the smooth endolining of the covered stent. However, although the
portosystemic gradient improved, no significant difference was noted in portal
venous stent velocities between the Viatorr and Wallstent groups (91.4 vs 89.4
cm/sec, respectively).
Many different protocols for sonographic TIPS evaluation and follow-up have
been described. The protocol used at our institution includes obtaining
intrastent velocities within the proximal and distal ends. Middle and peak
intrastent velocities have been described in the literature
[6] but are not routinely used
at our institution. The proximal and distal intrastent velocities serve as
primary data points for comparison of proximal versus distal velocities and
for future comparison to determine temporal changes in velocity. All have been
shown to be useful in evaluating for TIPS malfunction
[6]. Although the stent could
be visualized sonographically in all cases, after placement of the covered
stent, distal stent velocities could not be obtained in 9/19 (47%), which
suggests TIPS compromise including thrombosis. This represented a significant
difference when compared with the noncovered stent group, in which all distal
stents were visualized with flow (24/24, 100%).
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2D —45-year-old man with biopsy-proven nonalcoholic steatohepatitis.
Doppler sonogram obtained 12 days after TIPS procedure shows patency and
normalized velocities at hepatic venous end of shunt.
|
|
Further comparison of hepatic venous velocities between the Viatorr and
Wallstent groups elicited other differences. Comparing the initial normally
interpreted Viatorr and Wallstent groups, significant differences were
observed in hepatic venous stent velocity (132.4 cm/sec vs 116.0 cm/sec,
respectively). To our knowledge, this also has not been addressed in the
literature. It may also relate to the smooth endolining of the Viatorr stent
and should be investigated prospectively.
Lack of visualization of the hepatic venous stent occurred secondary to
documented thrombus in only one case. However, in this case
(Table 3) no portal venous flow
could be shown, and echogenic thrombus was evident on gray-scale imaging in
both the hepatic and venous systems. The patient died 2 days later, and
complete thrombosis of the hepatic and portal venous systems was shown at
autopsy.
Otal et al. [2] described a
high rate of lack of visualization of distal velocity in TIPS procedures using
covered stents at early sonographic follow-up
[2]. They described a
spontaneous-resolving acoustic barrier that prevented exploration of the shunt
lumen and noted that the hemodynamics in the TIPS with covered stents could be
recorded 1 week after the procedure. We sought to evaluate this finding more
thoroughly and also observed in our population that velocity in the distal end
of the stent was more easily obtainable at least 1 week after TIPS placement,
and overall interpretation was thus more adequate
(Fig. 2D). This delay in
baseline evaluation, which in our population occurred at 7 (n = 2),
12 (n = 1), and 14 (n = 1) days after TIPS, provided more
accurate distal stent velocities for subsequent follow-up and evaluation of
TIPS malfunction rather than the initial after-TIPS baseline.
Because many screening protocols depend on reliable velocity measurements,
nonvisualization of the distal covered stent at 24-hr follow-up remains
problematic. Because the 7- to 14-day follow-up sonography resulted in
significant improvement in visualization, we believe that unless clinical
suspicion of TIPS malfunction or procedural complication is present,
sonography should be delayed until 7–14 days after TIPS placement. If
immediate TIPS malfunction is suspected, sonography should remain the initial
examination because approximately 50% will be successful in demonstrating
patency. However, when sonography is nondiagnostic, further evaluation with CT
angiography or venography may be warranted despite the radiation and contrast
load [8].
The cause for the distal hepatic stent acoustic barrier is also a question
that is unresolved. Otal et al.
[2] explained that the acoustic
barrier may be related to microbullae embedded inside the ePTFE. As we know,
the ePTFE membrane is porous, and there is a significant amount of space
between the nodes and microfibrils of the polymer structure. Because the
acoustic barrier improves in a week, it may also be related to pockets of air
or gas between the ePTFE and the wire mesh. The device is restrained by an
ePTFE string wrap that also has a potential for carrying air when the device
is introduced into the delivery sheath. Other potentially slowly resolving
acoustic barriers may be related to the proliferation of myofibroblasts or
filling in spaces between the hepatic parenchyma and the ePTFE membrane with
aggregates of macrophages and fibrin as described pathologically after
explantation [12]. Further
evaluation with phantoms may be of use to further elucidate the cause of the
initial barrier.
Overall interpretation of the sonography findings is the most sensitive
criteria for TIPS malfunction. Kanterman et al.
[6] showed 92% sensitivity and
72% specificity for overall interpretation versus venography, which was more
sensitive than peak shunt velocity, change in peak shunt velocity, distal
shunt velocity, and main portal vein velocity for detecting shunt malfunction.
Because the integration of all sonography findings was most sensitive for TIPS
malfunction, overall interpretation was included as a primary outcome.
The limitations of our study are primarily related to the small population
size and the retrospective nature of the review. The limited sample of paired
examinations at follow-up also represents a significant limitation. However,
our results confirmed the existence of an acoustic barrier when using
ePTFE-covered stents and the differences between bare stents and covered
stents when used for TIPS. In addition, we contend that the optimal window to
sonographically evaluate and obtain a baseline study begins after 7 days.
Further confirmation of these findings with a prospective evaluation is
desirable, however.
In conclusion, according to the preliminary results of our study, the
sonographic evaluation of ePTFE-covered endoprostheses may be limited by a
resolving acoustic barrier. In situations of routine follow-up when procedural
complications and early TIPS malfunction are not suspected, a 7- to 14-day
evaluation may be more appropriate for early sonographic evaluation. These
results may be further validated through larger comparative studies of
sonographic evaluation of TIPS using varying time intervals.
References
- Coldwell DM, Ring EJ, Rees CR, et al. Multicenter investigation of
the role of transjugular intrahepatic portosystemic shunt in management of
portal hypertension. Radiology 1995;196
: 335-340[Abstract/Free Full Text]
- Otal P, Smayra T, Bureau C, et al. Preliminary results of a new
expanded-polytetrafluoroethylene-covered stent-graft for transjugular
intrahepatic portosystemic shunt procedures. AJR2002; 178:141
-147[Abstract/Free Full Text]
- Angermayr B, Cejna M, Koenig F, et al. Survival in patients
undergoing transjugular intrahepatic portosystemic shunt: ePTFE-covered
stentgrafts versus bare stents. Hepatology2003; 38:1043
-1050[CrossRef][Medline]
- Haskal ZJ. Improved patency of transjugular intrahepatic
portosystemic shunts in humans: creation and revision with PTFE stent-grafts.
Radiology 1999;213
: 759-766[Abstract/Free Full Text]
- Zizka J, Elias P, Krajina A, et al. Value of Doppler sonography in
revealing transjugular intrahepatic portosystemic shunt malfunction.
AJR 2000; 175:141
-148[Abstract/Free Full Text]
- Kanterman RY, Darcy MD, Middleton WD, et al. Doppler sonography
findings associated with transjugular intrahepatic portosystemic shunt
malfunction. AJR 1997;168
: 467-472[Abstract/Free Full Text]
- Lin EC, Middleton WD, Darcy MD, et al. Hemodynamics revealed by
Doppler sonography in patients who have undergone creation of transjugular
portosystemic shunts: comparison of 10- and 12-mm metallic stents.
AJR 1999; 172:1245
-1248[Abstract/Free Full Text]
- Haskal ZJ, Carroll JW, Jacobs JE, et al. Sonography of transjugular
intrahepatic portosystemic shunts: detection of elevated portosystemic
gradients and loss of shunt function. J Vasc Interv
Radiol 1997; 8:549
-556[Medline]
- Ong JP, Sands M, Younossi ZM. Transjugular intrahepatic
portosystemic shunts: a decade later. J Clin
Gastroenterol 2000; 30:14
-28[CrossRef][Medline]
- Feldstein V, LaBerge J. Sonography of TIPS. Tech Vasc
Interv Radiol 1998; 1:86
-93[CrossRef]
- Dodd GD, Zajko AB, Orons PD, et al. Detection of transjugular
intraportosystemic shunt dysfunction: value of duplex Doppler sonography.
AJR 1995; 164:1119
-1124[Abstract/Free Full Text]
- Cejna M, Peck-Radosavljevic M, Thurnher SA, et al. Creation of
transjugular intrahepatic portosystemic shunts with stent-grafts: initial
experiences with a polytetrafluoroethylene-covered nitinol endoprosthesis.
Radiology 2001;221
: 437-446[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
