| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Charité Hospital, Berlin, Germany.
2 Present address: Central Department of Diagnostic and Interventional
Radiology, Hospital Peine, Virchowstr. 8 h, Peine 31224, Germany.
Received December 6, 2004;
accepted after revision February 16, 2005.
Address correspondence to N. Hidajat.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty patients with portal hypertension due to liver cirrhosis were enrolled. In the irradiation group of 20 patients, the gamma radiation source, iridium-192, was introduced into the shunt within a special balloon catheter that allows the radionuclide to be centered within the shunt. A dose of 14 Gy in the shunt wall at a depth of 2 mm should be achieved. In the control group of 20 patients, a dummy source was used. Doppler sonography was performed immediately, 1 day, 1 week, 4 weeks, and 3 months after TIPS placement and then at an interval of 3 months during the first year. The primary end point of the study was the percentage of patients who developed significant shunt stenosis, defined as a reduction of maximum flow velocity below 50 cm/sec in the proximal part of the shunt 1 cm from the entry of the stent into the punctured portal vein branch. Fisher's exact test was used.
RESULTS. The TIPS procedure was technically successful in all patients. Seventeen patients in the irradiation group and 15 patients in the control group were followed up. Five patients (29.4%) in the irradiation group and 10 (66.7%) in the control group developed significant shunt stenosis during the first year after TIPS placement (p = 0.0392). The time until such stenosis occurred did not differ significantly between the two groups.
CONCLUSION. Our results suggest that brachytherapy can be useful in reducing the incidence of TIPS stenosis. A larger study with histopathologic analysis may be needed to confirm these findings.
Keywords: brachytherapy • liver disease • radiation therapy • shunts • sonography • stenosis • TIPS
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A problem in the treatment with TIPS is shunt stenosis, which can lead to shunt insufficiency, rebleeding, and recurrent ascites. The stenosis occurs predominantly at the hepatic vein end of the stent or in the native hepatic vein adjacent to the stent [2] and has been shown to be caused by pseudointimal hyperplasia, which is granulation tissue in association with myofibroblasts that increase in number over time [3, 4]. Intraluminal brachytherapy has been used successfully in the prevention of vascular restenosis of femoral and femoropopliteal arteries that is caused by neointimal hyperplasia, that is, proliferation of an extracellular matrix produced by stimulated smooth muscle cells. One possible problem of radiation therapy of arterial stenoses is early thrombosis [5, 6].
Although pseudointimal hyperplasia and neointimal hyperplasia are histologically different kinds of tissue, both are associated with proliferation of cells and cell products. The aim of this prospective randomized double-blind study was to examine whether intraluminal brachytherapy can prevent significant stenosis of TIPS.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Randomization
The patients were randomized into the irradiation group or the control
group. For randomization, the radiation therapist prepared 40 envelopes with a
sheet of paper inside. In 20 envelopes, the words "irradiation
group" and in the other 20, the words "control group" were
on the sheet. Every study patient drew an envelope. The result of
randomization was known only to the radiation therapist until the study was
terminated.
Procedure
To establish a portosystemic shunt, a puncture needle was advanced
transjugularly in a catheter through the inferior vena cava into the right
hepatic vein. An intrahepatic branch of the portal vein was punctured, and
self-expandable bare nitinol stents (Angiomed) with a diameter of 8 mm and a
length of 4 or 6 cm were implanted. The end of the stent was located in the
hepatic vein. The stents and the hepatic vein were dilated to 8 mm after
deployment. Pressure measurements were performed for the portal vein and right
atrium to ensure that the portosystemic pressure gradient was reduced to 10 to
15 mm Hg. When TIPS placement was completed, the 10-French angiographic sheath
was left with the tip in the superior vena cava. The guidewire was in the
superior mesenteric vein or the splenic vein. Then the patient was taken to
the radiation therapy room, and 5,000 IU of heparin was injected as a bolus
into the angiographic sheath. In addition, 2,500 IU of heparin in 1,000 mL of
a saline solution was infused slowly into the sheath.
For radiation therapy, the PARIS catheter system and the afterloader microSELECTRON-HDR (both from Nucletron) were used. The catheter system consisted of a double-lumen balloon catheter and a radiation sheath. On the balloon catheter was a radiopaque mark at a distance of approximately 12 mm to the tip of the catheter. The balloon diameter was 7 mm, the total length of the balloon was 10 cm, and the balloon was divided into 10 segments of 1 cm in length. The distal tip of the lowermost balloon segment corresponded approximately to the radiopaque mark. The radiation sheath was a closed-end tube that could be introduced into the balloon catheter. Because of the segmentation of the balloon, its inflation would center the catheter along the central axis of the vessel (Fig. 1).
|
Irradiation was performed in 5-mm steps. The radiation segment included the shunt from the entry of the stent into the punctured portal vein branch to the entry of the hepatic vein into the inferior vena cava plus 1 cm at both ends. The length of the radiation segment could be determined by the number of balloons that were in the segment (Fig. 2). A radiation dose of 14 Gy was aimed in a distance of 2 mm from the inflated balloon. Because the shunt diameter was 8 mm and the balloon diameter 7 mm, a dose of 14 Gy would occur in the shunt wall in a depth of 1–2 mm. The irradiation time per step depended on the activity of the nuclide. The total irradiation time was between 1 and 3 min. In patients in the control group, a dummy source was used instead of an active radionuclide. The time of "irradiation" with the dummy source was 5 sec per step.
|
|
|
End Points and Statistical Analysis
Primary end point—In both groups, the portion of patients
who developed significant shunt stenosis within the first year was determined.
According to published reports about radiation therapy of TIPS and arteries
[5–8],
the irradiation did not have a negative influence on patency. Therefore, we
started from a unilateral statistical test. The following hypotheses were
investigated:
 |
Because no data existed about radiation of TIPS immediately after placement, we estimated according to our own experiences [9] that 65% of the patients developed a significant shunt stenosis during the first year. According to Liermann et al. [5], none of the 18 patients who underwent irradiation developed a restenosis in the femoral artery. Teirstein et al. [7] found that 17% of the irradiated patients and 54% of the patients without irradiation developed restenosis in the coronary artery. On the basis of these data, we expected that 10–15% of the patients in the irradiation group and 65% of the patients in the control group would develop a significant shunt stenosis during the observation time. To ensure that the statistical comparison had a power of 90%, 20 patients were required in each group.
Secondary end point—In both groups, the time until
significant shunt stenosis occurred was determined. The following hypotheses
were investigated:
 |
The differences between the two groups were seen as statistically significant when the probability of error was less than 5%.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The irradiation group consisted of 20 patients (12 men, eight women; mean age, 54.1 ± 8.6 [SD] years; range, 35–68 years). Seventeen patients could be followed up. Three patients developed shunt occlusion due to thrombosis that was detected in the 1-week examination. Because this early shunt thrombosis was not caused by pseudointimal hyperplasia and the primary end point was significant shunt stenosis, these three patients were dropped from the study. In five of the remaining 17 patients, Doppler sonography detected a maximum velocity of less than 50 cm/sec at the 1-month (n = 2), 3-month (n = 2), and 9-month (n = 1) examinations. Twelve patients showed maximum velocity greater than 50 cm/sec at all Doppler sonography examinations (Fig. 3A).
The control group consisted of 20 patients (16 men, four women; mean age, 57.1 ± 9.6 years; range, 33–73 years). Fifteen patients could be followed up. Two patients died 4 and 5 days after TIPS placement, two did not come to the follow-up examinations after the 1-week examination, and one developed shunt occlusion due to thrombosis that was detected at the 1-week examination. These five patients were dropped from the study. In 10 of the remaining 15 patients, Doppler sonography detected maximum velocity within the stent of less than 50 cm/sec at the 1-month (n = 3), 3-month (n = 3), 6-month (n =2), 9-month (n = 1), and 12-month (n = 1) examinations. Five patients showed a maximum velocity greater than 50 cm/sec at all Doppler sonography examinations (Fig. 3B).
These results mean that the rate of significant shunt stenosis in the irradiation group (5/17 patients) was lower than that in the control group (10/15 patients) (29.4% vs 66.7%; p = 0.0392) during the observation time. No significant difference was seen between the irradiation group and the control group in the time until significant shunt stenosis was detected in the five and 10 patients, respectively. The difference in the rate of early shunt thrombosis between the irradiation group (3/20 patients) and the control group (1/18 patients) was not significant (15.0% vs 5.6%; p = 0.3436).
During the observation time, no bleeding episodes occurred. In all patients with refractory ascites as an indication for TIPS placement, the ascites disappeared, but some patients with significant shunt stenosis in the control group experienced recurrent ascites.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The positive effect of irradiation in the prevention of granulation tissue in the skin has been known for many years [12]. Some authors report endovascular radiation therapy for the prevention of pseudointimal granulation tissue in arteries after angioplasty and in TIPS. In the study of Liermann et al. [5], 18 patients with restenosis of the femoral artery after angioplasty were treated by such radiation therapy. None of the patients developed restenosis in the irradiated vessel segments, but in two of the patients stenosis occurred below these segments. Krueger et al. [6] showed that endovascular radiation therapy can reduce the degree of stenosis after de novo femoropopliteal stenoses. The antiproliferative effect of intracoronary radiation therapy in patients with coronary restenosis was shown by Teirstein et al. [7].
Intraluminal brachytherapy of TIPS with 192Ir has been performed in humans after TIPS revision. Pokrajac et al. [13] and Dvorak et al. [14] reported on five and six patients, respectively. The endovascular radiation therapy was safe and feasible in both studies. A dose of 12 or 15 Gy was applied at a distance of 3 or 5 mm from the radiation source. However, the stent diameter was 10 mm and the radiation source was not centered in the lumen of the stent [13, 14]. This means that the dose distribution at the stent wall could not be homogeneous and that the dose was much lower than 12 Gy in some parts of the shunt wall and much greater than 15 Gy in other parts.
Recently, Hausegger et al. [8] reported endovascular TIPS irradiation in five pigs. The authors documented a lesser degree of endothelialization of the inner surface of the stentgraft in the irradiated group compared with the control group. However, no significant differences in patency could be observed between the two groups. In six swine with normal portal pressures, Lessie et al. [15] irradiated TIPS with a dose of 15.2 Gy of phosphorus-32 delivered at the time of TIPS placement. The degree of pseudointimal hyperplasia was not reduced compared with a control group.
To our knowledge, our study is the first randomized prospective and double-blind study of intraluminal brachytherapy of TIPS in human patients immediately after TIPS placement. Moreover, we used a centering balloon catheter, which is obligatory to secure a homogeneous dose distribution in the vessel wall.
Because the maximum balloon diameter was 7 mm, we decided to implant stents with a diameter of 8 mm. This secured a small gap between the balloon wall and the stent wall and continuous blood flow when the balloon was inflated, and thus diminished the risk of thrombosis. On the other hand, if a larger stent were used, the radionuclide would deviate too much from the center of the stent and the dose distribution in the stent wall would be too inhomogeneous. We did not use beta irradiation because the maximum range is too low for such a great lumen diameter.
According to most articles, TIPS stenosis is most frequently related to pseudointimal hyperplasia in the hepatic vein end of the stent or in the native hepatic vein adjacent to the stent [2, 16]. However, Saxon et al. [17] reported that abnormalities of the parenchymal tract were more often correlated with recurrent variceal bleeding or ascites than were hepatic vein stenoses. Therefore, brachytherapy must include the whole shunt, as in our study. We could show that intraluminal brachytherapy as described in this article can reduce the incidence of significant shunt stenosis. However, we suggest that in some patients the irradiation may not stop the development of stenosing granulation tissue. Thus, in some patients, significant shunt stenosis develops as early as in patients without irradiation. This finding may agree with that of Lessie et al. [15] that irradiation may not prevent or reduce pseudointimal hyperplasia in all cases.
According to the previously mentioned studies, intraluminal brachytherapy of TIPS can be safe and feasible. The numbers of deaths (2/40 patients) and of patients with early shunt thrombosis (4/38 patients) during the first week in our study are not significantly different from those in patients in our department reported previously [9]. The irradiation of TIPS cannot be seen as a risk of thrombosis because the rate of early shunt thrombosis was not significantly different between the irradiation and the control groups. However, we must be cautious with this result because of the small number of patients.
In the patients with early shunt thrombosis and in some of the patients with significant shunt stenosis detected on Doppler sonography, TIPS revision was performed and confirmed the shunt insufficiency. In other patients, we postponed the TIPS revision because of the patient's poor condition or decreasing liver function. In all patients with stent revisions and stenosis in the irradiation group, the stenosis was located only at the region of the hepatic vein end of the stent. In some of the patients with stent revisions and stenosis in the control group, the stenosis could also be seen in the parenchymal tract. Probably the injury of the hepatic vein wall by balloon angioplasty and stent leads to a tendency to a more excessive cell proliferation, and a higher radiation dose is necessary to stop the development of significant pseudointimal hyperplasia. Possibly, significant stenosis at the hepatic vein end of the stent could be overcome by extending the stent into the hepatocaval junction at the time of TIPS placement, as found by Clark et al. [18].
Many studies have confirmed that Doppler sonography can be used for follow-up. A variety of hemodynamic parameters—including flow velocity in the main portal vein, maximum or minimum flow velocity in the stent, flow direction in the intrahepatic portal and hepatic veins, and temporal changes in the velocity and blood flow direction—can be used [19]. Angiography and measurement of the portosystemic gradient is no longer a standard follow-up examination because it has been shown that Doppler sonography is accurate in the noninvasive assessment of shunt stenosis [2, 20].
Our experience agrees with that of Chong et al. [20] that significant shunt stenosis can be predicted when the maximum flow velocity is less than 50 cm/sec in the proximal part of the shunt about 1 cm from the entry of the stent into the portal vein branch [9]. Power Doppler sonography with or without contrast enhancement and CT angiography have been shown to be helpful in patients who previously underwent unsatisfactory sonography [21, 22]. However, these methods were not necessary in the patients in our study.
Several histologic studies of TIPS helped us to define the time at which a shunt stenosis in a TIPS can be assumed to be caused by pseudointimal hyperplasia. According to LaBerge et al. [3], a single layer of endothelial cells can begin to line the shunt surface 4 days after TIPS placement. Two to three weeks after TIPS placement, the stent is covered by a smooth pseudointima of granulation tissue that extends with time as the myofibroblasts increases in number and the collagen fibers thickens [3, 4]. After this time, pseudointimal hyperplasia is the main cause of TIPS stenosis or occlusion [4]. This finding agrees with that of Terayama et al. [23] in stenosed TIPS 49 and 293 days after placement. On the basis of these findings, we decided to drop patients from our study who showed shunt occlusion at the 1-week examination because such occlusion is caused by thrombosis and not by pseudointimal hyperplasia.
Because some patients dropped out of the study (three patients in the irradiation group and five in the control group), the number of patients to be followed up became smaller. This small number may be a limitation of the study. Also, we did not follow up the patients for longer than 12 months.
In recently published articles, the use of an expanded polytetrafluoroethylene (ePTFE)-covered stent-graft is recommended for TIPS. This stent can solve pseudointimal hyperplasia in the stent track [24]. To prevent hepatic or portal vein stenosis and to improve the patency rate compared with conventional stents, the ePTFE stent should be placed up to the inferior vena cava [25]. In our opinion, a major disadvantage of this covered stent is the high cost. The material cost for intraluminal brachytherapy can be expected to be much lower.
We conclude that intraluminal brachytherapy can decrease the incidence of significant stenosis of TIPS. The number of revisions during the first year can be expected to reduce. The prevention or reduction of pseudointimal hyperplasia should be confirmed by histopathologic studies. Further and larger studies are needed to assess survival rates and to compare primary implantation of ePTFE stent-grafts for patency and financial costs.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
