| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
i
ka1
1ka1ková1
lek2
afka2 Va
ásek2
3
1
Department of Diagnostic Radiology, Charles University Hospital,
Sokolská 408, CZ-500 05 Hradec
Králové, Czech
Republic.
2
Department of Internal Medicine, Charles University Hospital, CZ-500 05 Hradec
Králové, Czech
Republic.
3
Department of Biophysics, Medical Faculty, Charles University, CZ-500 38
Hradec
Králové, Czech
Republic.
Received October 4, 1999;
accepted after revision December 7, 1999.
Supported by grant 4109-3 of the Internal Grant Agency, Ministry of Health,
Czech Republic. ika.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. During a 5-year period, 1192 Doppler examinations were performed in 216 patients with TIPS. No regular follow-up shunt venography was performed. Doppler examinations were retrospectively compared with the results of shunt revisions. Sonograms with negative findings were compared with the patients' clinical status so that the number of false-negative sonographic findings leading to an episode of shunt failure (recurrence of gastrointestinal bleeding or ascites) could be ascertained. Sonographic parameters assessed included diameter, velocity, flow volume, and congestion index of the portal vein; and shunt velocities.
RESULTS. Doppler sonography revealed shunt occlusion in 25 of 26
angiographically proven cases (sensitivity, 96%). The combination of velocity
criteria (peak intrashunt velocity 
250 cm/sec, maximum velocity in the
portal third of the shunt 
50 cm/sec, or maximum portal vein velocity less
than or equal to two thirds of the baseline value) revealed shunt stenosis in
103 of 110 cases (sensitivity, 94%). Doppler sonography missed a significant
shunt stenosis that led to an episode of gastrointestinal bleeding or ascites
recurrence in only seven cases. The congestion index of the portal vein showed
significant differences between patent and malfunctioning shunts
(p<0.001).
CONCLUSION. Doppler sonography is an effective primary imaging method for long-term follow-up of patients with TIPS.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Because of its low cost and its noninvasive nature, Doppler sonography is generally accepted as the primary examination for detecting shunt failure. Using various definitions of shunt patency and different sonographic criteria, most authors proved its efficacy and accuracy in shunt compromise detection [8,9,10,11,12,13,14,15,16,17]. However, two recent reports suggest that Doppler sonography is not a sensitive test for predicting the presence of a hemodynamically significant shunt stenosis and that relying solely on noninvasive tests is not advised for TIPS follow-up [18,19].
The aim of this study was to evaluate the long-term clinical efficacy of Doppler sonography in predicting shunt failure. We also attempted to assess the hemodynamic changes related to TIPS placement and to give more precision to selected Doppler criteria for identification of shunt malfunction.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Doppler sonography was the sole imaging method for regular follow-up of patients with TIPS. Shunt venography was reserved exclusively for therapeutic interventions (i.e., angioplasty or stent placement). Our follow-up protocol consisted of sonographic examinations at 1-3 days, at 7-10 days, and then at 3-month intervals after TIPS creation. Starting from the second year of follow-up, the intervals were prolonged to 6 months. This regular protocol was applied when no signs of shunt malfunction existed. When significant shunt stenosis or occlusion was revealed on sonography or was suspected from the clinical status or both, the patient was referred for angiographic shunt revision. After the revision, the patient reentered the same follow-up protocol from the start except for the examination at 7-10 days, which was carried out only after TIPS insertion. The sonographic findings obtained immediately after shunt creation or revision (baseline study) served for comparison with the results of subsequent examinations.
All patients were sonographically examined with an Ultramark 9 HDI unit (Advanced Technology Laboratories, Bothell, WA). For morphologic assessment, a 4-2-MHz broadband curved linear transducer was used. For Doppler evaluation, a 3-2-MHz broadband phased array transducer was used.
Immediately after each sonographic examination, the data were recorded into a computer protocol and were subsequently printed and sent to a gastroenterologist. The protocol included the data about the cross-sectional anteroposterior diameter and maximum flow velocity (Vmax) of the portal vein measured halfway between the confluence and the portal bifurcation (Fig. 1). Shunt velocity was routinely measured in the portal third of the TIPS approximately 2 cm downstream from the junction of the TIPS with the portal vein (Fig. 2). We tried to avoid measurements in the entry segment of the stent located directly in the portal vein. In addition to the maximum shunt velocity measured in the portal third of the TIPS, we recorded peak intrashunt velocity irrespective of the measurement site along the entire course of the acoustically accessible shunt and the draining hepatic vein (Fig. 3A,3B,3C). For all velocity measurements, subcostal or intercostal approaches were used to avoid measurements with Doppler angles exceeding 60°. No measurements were made at angles greater than 70°. Color Doppler sonography was routinely used as a "road map" for spectral velocity measurements.
|
|
|
|
|
The shunt was sonographically considered to be occluded when no Doppler signal was seen in the stent. A hemodynamically significant stenosis was assumed in those shunts in which at least one of the following criteria was fulfilled: peak intrashunt velocity was greater than or equal to 250 cm/sec (Fig. 3A,3B,3C), maximum velocity in the portal third of the shunt was less than or equal to 50 cm/sec (Fig. 2), or maximum portal vein velocity was less than or equal to two thirds (66.7%) of the baseline value (Fig. 1). We determined these criteria on the basis of the data available in the literature at that time [8, 9, 20, 21] and on the basis of our results of seven correlations between sonographic and positive angiographic findings obtained in 23 patients who were examined before January 1994 and were not included in the study.
During the 5-year period, 1192 Doppler examinations in 216 patients with TIPS were performed. All shunts that were stenosed or occluded on sonography were referred for angiographic revision; in cases of discrepancy between negative sonographic and positive clinical findings (recurrence of gastrointestinal bleeding or ascites), the shunt revision was indicated by a gastroenterologist. In this manner, 143 angiographic shunt revisions in 86 patients were performed (range, 1-7 per patient). For purposes of this study, we considered each series of follow-up sonographic examinations starting with the baseline study after shunt creation or revision and lasting until the next possible revision to be a single follow-up period (unit). Thus, a total of 359 sonographic follow-up periods were available for retrospective evaluation (216 series after shunt creation and 143 series after shunt revision). The mean duration of the follow-up period was 10.6 months (range, 1 day to 52 months). Sonographic follow-up was discontinued before February 1999 because of death in 47 patients, transplantation in 12 patients, and loss to follow-up of 16 patients. Additionally, sonographic data obtained shortly (<10 days) before the TIPS creation were available in 144 patients.
TIPS were catheterized via an internal jugular approach. Real-time hepatic and portal vein pressures were recorded routinely. The shunt was defined as stenosed if a portosystemic gradient of more than 12 mm Hg was found. Shunt occlusion was defined by the complete absence of flow within its lumen. Malfunctioning (i.e., stenosed or occluded) shunts were treated by balloon angioplasty or placement of an additional stent.
In this manner we obtained the data set of sonographic follow-up results for retrospective correlation with the clinical outcomes of 216 patients. We evaluated the capability of our sonographic follow-up protocol to protect the patient from complications of portal hypertension recurrence. On the basis of the clinical data gathered over a 5-year period, the percentage of false-negative sonographic findings leading to an episode of shunt failure (recurrence of gastrointestinal bleeding or ascites) was ascertained. The set of 143 angiographic shunt revisions permitted us to evaluate the value of our previously mentioned sonographic criteria for depicting shunt stenosis or occlusion. To accentuate the differences between normal and malfunctioning shunts, we retrospectively evaluated two additional hemodynamic criteria. First, the flow volume in the portal vein was calculated by multiplying the cross-sectional area and 0.57 Vmax of the portal vein. The coefficient 0.57 derives the mean velocity from Vmax in the portal vein [22]. Second, the congestion index of the portal vein, which was defined by Moriyasu et al. [22] as a ratio between the cross-sectional area and the mean flow velocity of the portal vein, was calculated. Moreover, the sonographic data obtained before TIPS creation in 144 patients served for a close characterization of sonographically evident hemodynamic changes related to TIPS placement, which we believe to be important for understanding the value of the initial baseline sonographic study.
Sonographic data from before and after TIPS placement and from baseline and prerevision examinations were compared using a two-tailed t test for paired data. Baseline and prerevision velocities and flow volumes in the portal vein were compared between the groups of stenosed and occluded shunts using a two-sample t test. Significance was set at the 0.05 level in all cases. Data are presented as mean ± 1 SD.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Diagnostic Value of Sonographic Criteria
Of the total of 143 shunt revisions, angiography revealed 26 shunt
occlusions, 110 shunt stenoses, and seven patent shunts. Among the seven
patent shunts, four were incorrectly identified as stenotic on sonography;
three shunts had true-negative sonographic results (one shunt was indicated
for revision on the clinical suspicion of shunt malfunction despite negative
sonographic findings; and two shunts were indicated for placement of a
reducing stent because of severe portosystemic encephalopathy, and both were
patent on both sonography and angiography).
Of 136 angiographic revisions of malfunctioning shunts, 129 shunt revisions (95%) were correctly indicated by true-positive sonographic findings. In only seven cases (5%) was the indication for shunt revision based solely on the clinical findings, without supporting sonographic findings, which were false-negative.
Baseline parameter values obtained from patent shunts are compared with those of malfunctioning (i.e., stenosed plus occluded) shunts in Table 2. The values obtained separately from the groups of stenosed or occluded shunts are listed in Table 3.
|
|
Detection of Shunt Occlusion
During the period of the study, 25 shunts showed no Doppler signal on
sonography and were qualified as occluded. These occlusions were confirmed in
all cases by shunt venography. One additional case of shunt occlusion was
revealed on venography: this occlusion was described as critically stenotic on
sonography 1 day before shunt revision, with a peak shunt velocity 240 cm/sec
and maximum velocity in the portal third of the shunt of 30 cm/sec. No patent
shunts were described as occluded on sonography. The sensitivity for shunt
occlusion detection was 96.2%, and specificity was 100%. Positive predictive
value for shunt occlusion was 100%, positive predictive value for shunt
patency was 99.2%, and accuracy was 99.3%.
Detection of Shunt Stenosis
A hemodynamically significant stenosis was assumed when at least one of the
following criteria was positive: peak intrashunt velocity greater than or
equal to 250 cm/sec, maximum velocity in the portal third of the shunt less
than or equal to 50 cm/sec, or maximum portal vein velocity less than or equal
to two thirds (66.7%) of the baseline value. Using this definition, the
overall sensitivity for shunt stenosis detection was 93.6%, and the positive
predictive value was 96.3%.
Intrashunt Velocities
When comparing baseline values with the values obtained from stenosed
shunts, the mean peak intrashunt velocity increased by 64.9 cm/sec (p
< 0.001) and the mean Vmax in the portal third of shunt declined
by 21.5 cm/sec (p < 0.01)
(Table 3). In the depiction of
shunt stenosis, peak intrashunt velocity greater than or equal to 250 cm/sec
yielded a sensitivity of 50.9%. The velocity in the portal third of the shunt
less than or equal to 50 cm/sec yielded a sensitivity of 33.6%.
Portal Vein Parameters
Mean Vmax and flow volume in the portal vein declined
significantly in compromised shunts by 14.1 cm/sec and 479 ml/min,
respectively (p < 0.001 for both parameters)
(Table 2). The decline in these
two parameters was more pronounced in the group of occluded shunts (p
< 0.001 for Vmax and p < 0.01 for flow volume) than
in the group of stenosed shunts (p < 0.01 for both parameters)
(Table 3). Using a 33% decrease
as a cutoff, the sensitivity for shunt stenosis depiction was higher in portal
vein velocity measurements (51.1%) than in flow volume measurements
(31.8%).
Statistically significant differences between baseline and prerevision values of the congestion index and diameter of the portal vein were found (p < 0.001) (Table 2).
Clinical Efficacy
Among 216 patients who were treated with TIPS in the course of our 5-year
study, 130 patients had primarily patent shunts and no clinical evidence of
shunt failure during the sonographic follow-up. Among the remaining 86
patients who had at least one shunt reintervention, 57 patients had positive
sonographic findings and no clinical signs of shunt failure, 22 patients had
positive sonographic and positive clinical findings, and seven patients had
negative sonographic and positive clinical findings.
A total of 359 sonographically documented interventions (216 primary TIPS procedures plus 143 TIPS reinterventions) were performed. Consequently, our complete material contained 359 baseline sonographic studies succeeded by a total of 833 follow-up sonograms. Each baseline study and its subsequent sonographic examinations thus formed a single shunt follow-up period lasting until the next intervention was performed.
Within the 359 shunt follow-up periods, over a 5-year period we noted seven falsenegative sonographic examinations leading to a clinical complication caused by shunt stenosis. Those clinical complications and indications for shunt revision were variceal bleeding recurrence in three patients, portal gastropathy bleeding in one patient, portal gastropathy bleeding with high portosystemic gradient and patent shunt at angiography in one patient (solved by creation of a second TIPS), and non-resolving ascites after TIPS insertion in two patients (caused by incompletely dilated patent shunt in one patient and mural midshunt thrombosis in the other). From the clinical point of view, Doppler sonography failed to depict a significant shunt stenosis leading to an episode of gastrointestinal bleeding or ascites in seven (2%) of 359 follow-up periods.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Diagnostic Value of Sonographic Criteria
The major aim of our study was to evaluate the clinical efficacy of Doppler
sonography in predicting shunt failure. Because the study was primarily based
on noninvasive sonographic evaluation of the functional status of TIPS, we
were able to compare with the gold standard of angiography only those shunts
that were either qualified as stenotic or occluded on sonography or those for
which revisions were, in rare cases (5%), indicated by a gastroenterologist
because of a strong clinical suspicion of shunt malfunction despite negative
sonographic findings. We calculated sensitivities and positive predictive
values of major sonographic parameters describing the shunt function. Because
we did not perform any prospective shunt venography, we had no large set of
data with angiographic findings from normal TIPS. That is why we did not
calculate the specificities and negative predictive values for shunt stenosis
detection. As a counterpart to these statistical descriptors, we evaluated the
longterm ability of Doppler sonography to protect a patient with TIPS from
severe clinical complications accompanying an episode of portal hypertension
recurrence (i.e., gastrointestinal bleeding or ascites).
Mean Vmax measured in the portal vein was significantly lower in patients with malfunctioning shunts than their baseline values (Vmax decreased by 32% from 43.1 cm/sec to 29.0 cm/sec; p < 0.001). This corresponds well with almost doubling the portosystemic gradient (from 9.4 to 18.1 mm Hg) in the group of malfunctioning shunts compared with their baseline values and also corresponds with the results of other researchers [15, 16, 27]. Mean VMax in the portal vein for the group of well-functioning shunts ranged from 41 to 47 cm/sec in five other studies [11, 15, 16, 21, 27] as well as in ours (43.1 cm/sec). A broader spectrum of mean Vmax in the portal vein (24-33 cm/sec) belongs to the groups of malfunctioning shunts reported by these authors; our velocity value of 29.0 cm/sec does not reach outside this range, either.
The individual and interindividual differences of Vmax in the portal vein make the choice of a single threshold value discriminating patent from malfunctioning shunts difficult. Although Haskal et al. [16] reported a threshold of 40 cm/sec and predicted shunt failure with 84% sensitivity and 54% specificity, Kanterman et al. [15] obtained almost identical sensitivity of 82% with a threshold of 30 cm/sec (specificity, 77%). For more precise assessment of shunt function, a choice of relative temporal change from the baseline portal vein velocity seems to be more appropriate. Using a 20% decrease as a cutoff, Kanterman et al. obtained a sensitivity of 78% and specificity of 75%. Using a 33% cutoff value, their sensitivity declined to 67%, with specificity unchanged. In our study, the 33% cutoff value yielded the sensitivity of 51.1% and positive predictive value of 100%.
Changes in shunt and portal venous flow volumes were inversely correlated (p < 0.001) with changes in the portosystemic gradient in the study of Lafortune et al. [23]. In our study, we found a significant difference between the groups of baseline and prerevision portal venous flow volumes (p < 0.001) (Fig. 4). However, much easier measurement of temporal change of Vmax in the portal vein achieved similar results at the same probability level (Fig. 5). Moreover, the measurements of Vmax in the portal vein achieved greater sensitivity (51.1%) than the flow volume measurements (31.8%). Thus, we conclude that the simple measurement of Vmax in the portal vein is not only easier, but also a more exact predictor of shunt compromise than more complicated portal venous flow volume assessment.
|
|
The congestion index of the portal vein was defined by Moriyasu et al. [22] as the ratio between the cross-sectional area and the mean flow velocity of the portal vein. The role of the congestion index is to numerically emphasize the characteristic signs of portal hypertension (i.e., decrease of blood flow velocity in an enlarged portal vein). A statistically significant difference between congestion indexes from the healthy subject group and from patients with liver cirrhosis was reported by those authors (0.070 ± 0.029 cm · sec versus 0.171 ± 0.075 cm · sec; p < 0.001). There were only sporadic reports about the use of this parameter in the evaluation of patients with liver cirrhosis in the 1990s. Interestingly, two recent articles report the usefulness of the congestion index for detecting portal hypertension. Merkel et al. [28] reported a significant linear correlation between hepatic venous pressure gradient and congestion index. Haag et al. [26] found a significantly increased congestion index in the study of 375 patients with portal hypertension when compared with healthy individuals (+185%; p < 0.001); if the congestion index exceeded 0.10 cm · sec, portal hypertension was diagnosed with a 95% sensitivity and specificity in the study by Haag et al. However, no data concerning changes in the congestion index after the TIPS procedure or during the follow-up of patients with TIPS have been published to date. To the best of our knowledge, ours is the first study describing the evolution of the congestion index before and after TIPS creation and during the long-term follow-up of patients with TIPS. We found a significant decrease of the congestion index immediately after shunt creation (from 0.143 ± 0.062 cm · sec to 0.060 ± 0.038 cm · sec; p < 0.001) (Fig. 6). We also found a significant increase of this parameter in patients with malfunctioning shunts when compared with their baseline values (from 0.063 ± 0.030 cm · sec to 0.095 ± 0.038 cm · sec; p < 0.001) (Fig. 7). We suggest that the congestion index may be a helpful additional criterion when evaluating TIPS for possible malfunction (it is already a criterion in the case of portal hypertension depiction). The role of the congestion index may even increase with the introduction of polytetrafluoroethylene-covered stent-grafts, which are not well penetrated by ultrasound waves and may hinder direct shunt velocity measurements. A similar effect may be anticipated on contrast-enhanced Doppler sonography [29, 30]—in this case, however, at the cost of increased expenditure.
|
|
Considerable attention has been paid by various authors to shunt velocity measurements. Their results cannot be easily compared because of their different definitions of a malfunctioning shunt and different velocity thresholds or measurement sites. Almost all authors reported high mean peak velocities obtained from well-functioning shunts, generally ranging from 65 to 220 cm/sec [8, 9, 11, 15, 16, 21], and in the study of Feldstein et al. [13], even ranging from 50 to 270 cm/sec.
Other investigators have tried to establish a lower limit of normal for shunt velocity. Chong et al. [8] defined a threshold of 50 cm/sec from the portal part of the shunt to have a sensitivity of 100% and a specificity of 93% for shunt stenosis detection, but that study included only eight stenosed shunts. In a group of 25 stenoses, Feldstein et al. [13] reported a 78% sensitivity and 99% specificity using this criterion. Haskal et al. [16] described a 61% sensitivity and 74% specificity of this cutoff value in a group of 37 stenosed shunts. Foshager et al. [11] found a threshold velocity value of 60 cm/sec to have a sensitivity of 100%. On the contrary, Dodd et al. [12] reported that the cutoff value of 60 cm/sec failed to show stenoses in 14 of 15 patients in their study (sensitivity, 7%) but suggested that a temporal change in peak stent velocity exceeding 50 cm/sec from the baseline value reached a 93% sensitivity and 77% specificity. Kanterman et al. [15] also found the cutoff value of 50-60 cm/sec too low and suggested the lower limit for peak shunt velocity be 90 cm/sec. These apparent differences can be explained by different definitions of shunt stenosis or Doppler measurement sites. Also, no general agreement exists concerning the upper velocity limit for normal shunts, usually proposed to be approximately 185-220 cm/sec [9, 15, 21]. However, Foshager et al. reported the upper range of peak velocities in well-functioning shunts to be 300 cm/sec in the first 2 months after TIPS creation.
When we were defining and refining our shunt velocity criteria in the early 1990s, only a limited knowledge existed about the threshold values distinguishing normal from abnormal shunts. We, as well as several other authors, started from two basic hemodynamic assumptions. First, if a shunt contains a significant stenosis, a focal elevation of shunt velocity at the site of stenosis must be present (Fig. 3A,3B,3C). Second, if the stenosis cannot be directly insonated because of unfavorable acoustic conditions (frequently deep in the outflow tract), a prestenotic velocity decline should be noted (Fig. 2). That is why we always not only recorded Vmax in the intraparenchymal portal third of the shunt (easily visualized in all cases) but also tried to examine the entire length of the shunt looking for peak velocity denoting the site of stenosis. If we found a focal elevation of shunt velocity greater than 250 cm/sec in any portion of the shunt, we called this location significantly stenosed and indicated shunt revision. However, of 110 shunts with angiographically proven stenoses, we did not find focal peak velocity exceeding 150 cm/sec in 14 shunts; 19 shunts did not show peak velocity exceeding 200 cm/sec (in all these cases, the stenosis was located in the outflow tract or in the hepatic vein—i.e., in the parts of the shunt where an adequate Doppler signal can be difficult to obtain). In such cases, the true site of stenosis was probably not directly accessible by Doppler sonography because of the limited acoustic conditions, and the diagnosis of shunt stenosis was made by means of decreased Vmax in the portal third of the shunt or decreased Vmax in the portal vein. This also explains the relatively low sensitivity (50.9%) of the isolated peak velocity threshold of 250 cm/sec in our study. Similarly, isolated cutoff values for Vmax in the portal third of the shunt also did not show high sensitivities (Vmax <50 cm/sec, sensitivity of 33.6%; Vmax <60 cm/sec, sensitivity of 57.3%) and were not so high as reported in some other studies; relying solely on this criterion would miss each stenosis located directly in the portal third of the shunt (7% in our study).
Even though many studies confirmed that Doppler sonography is an effective screening tool for the evaluation of the functional status of a TIPS, with good ability for detecting shunt malfunction [8,9,10,11,12,13,14,15,16,17, 20, 21, 23, 27, 29,30,31,32], two recent studies did not find it to be a sensitive test in revealing the presence of hemodynamically significant stenosis. Murphy et al. [18] reported that maximum shunt velocity of 60 cm/sec or less was 93% specific but only 25% sensitive for detecting shunt stenosis. Owens et al. [19] detected shunt malfunction in only 11 of 31 cases (sensitivity, 35%; specificity, 83%). Finally, both studies do not advise relying solely on sonography for follow-up of patients with TIPS and recommend invasive portal pressure measurements for regular follow-up. In our opinion, the diagnosis of shunt malfunction needs a multifactorial approach, which was not consistently applied in these two studies. If we combine two or even three velocity criteria, the sensitivity rises considerably. Similarly, a temporal change of various parameter values (when compared with the baseline) is a helpful sign of TIPS malfunction. Our results are also in agreement with the results of five other studies that reported considerably higher sensitivity when evaluating either a combination or a temporal change of parameters [11, 12, 15, 16, 23].
Clinical Efficacy
Although our study was neither prospective nor double-blinded, it was a
5-year study in 216 patients with TIPS and comprised almost 1200 shunt
sonographic examinations. Of 136 angiographic revisions of malfunctioning
shunts, 129 revisions (95%) were indicated by sonography and seven revisions
(5%) were indicated by strong clinical suspicion of shunt failure. We did not
perform any prospective shunt venography. In our study, Doppler sonography
yielded a sensitivity of 96.2% for the detection of shunt occlusion. When
using a combination of three velocity criteria (peak intrashunt velocity
250 cm/sec, maximum velocity in the portal third of the shunt 50
cm/sec, maximum portal vein velocity less than or equal to two thirds of the
baseline value), we achieved a 93.6% sensitivity for shunt stenosis detection.
Of 359 baseline studies and their follow-up periods, sonography missed a
significant shunt stenosis leading to an episode of gastrointestinal bleeding
or ascites recurrence in only seven cases (2%).
In conclusion, our results suggest that Doppler sonography is a reliable noninvasive imaging technique for evaluating the functional status of TIPS. Timely detection of shunt malfunction is desirable to prevent complications of portal hypertension recurrence, particularly variceal bleeding. This method proved to have a high sensitivity for detecting shunt occlusion and significant shunt stenosis (when using either a combination or temporal change of multiple velocity criteria). Such an approach allows Doppler sonography to be used as a primary tool for the long-term follow-up of patients with TIPS. Temporal change of selected additional criteria (e.g., congestion index) may be a helpful sign of imminent shunt failure. Invasive shunt venography should, in our opinion, be reserved for therapeutic purposes.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
P, ika J.
Portal system, hepatic veins. In:
Eliá P,
ika J, eds.
Doppler ultrasonography. Hradec
Králové, Czech
Republic: Nucleus, 1998:209
-240
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Itkin, A. Patel, S. O. Trerotola, and H. I. Litt MRI of TIPS with Covered Stent-Grafts: In Vitro Analysis Using a Flow Phantom and Initial Clinical Experience Am. J. Roentgenol., June 1, 2009; 192(6): W317 - W320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Hoppe, S. L. Wang, and B. D. Petersen Intravascular US-guided Direct Intrahepatic Portocaval Shunt with an Expanded Polytetrafluoroethylene-covered Stent-Graft Radiology, January 1, 2008; 246(1): 306 - 314. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Lake, M. Guimaraes, S. Ackerman, C. Hannegan, C. Schonholz, J. B. Selby, and R. Uflacker Comparative Results of Doppler Sonography After TIPS Using Covered and Bare Stents. Am. J. Roentgenol., April 1, 2006; 186(4): 1138 - 1143. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Vignali, I. Bargellini, M. Grosso, G. Passalacqua, F. Maglione, F. Pedrazzini, P. Filauri, R. Niola, R. Cioni, and P. Petruzzi TIPS with Expanded Polytetrafluoroethylene-Covered Stent: Results of an Italian Multicenter Study Am. J. Roentgenol., August 1, 2005; 185(2): 472 - 480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. G. Sheiman, T. Vrachliotis, D. P. Brophy, and B. J. Ransil Transmitted Cardiac Pulsations as an Indicator of Transjugular Intrahepatic Portosystemic Shunt Function: Initial Observations Radiology, July 1, 2002; 224(1): 225 - 230. [Abstract] [Full Text]
|
|
|
|
|
|
M. Cejna, M. Peck-Radosavljevic, S. A. Thurnher, K. Hittmair, M. Schoder, and J. Lammer Creation of Transjugular Intrahepatic Portosystemic Shunts with Stent-Grafts: Initial Experiences with a Polytetrafluoroethylene-covered Nitinol Endoprosthesis Radiology, November 1, 2001; 221(2): 437 - 446. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
