Transjugular intrahepatic portosystemic shunt (TIPS) have been shown to be an efficient portal-systemic derivative treatment for Budd-Chiari syndrome (BCS) patients uncontrolled by medical therapy. However, the main drawback of TIPS for this condition is a very high rate of shunt dysfunction. Recently, polytetrafluoroethylene (PTFE)-covered stents have been shown to reduce the incidence of TIPS dysfunction in patients with cirrhosis. The aim of the study was to assess the incidence of TIPS dysfunction in 2 cohorts of BCS patients treated with bare or PTFE-covered stents. The study included 25 TIPS procedures (16 bare stents and 9 covered stents) with a median follow-up period of 20.4 months (range, 3.9-124.8). Fourteen of 16 patients (87%) receiving bare stents had TIPS dysfunction compared to 3 of the 9 patients (33%) receiving PTFE-covered stents (P = .005). The actuarial rates of primary patency in the bare-stent group were 19% at 1 year compared with 67% in the PTFE-covered stent group (P = .02; log-rank test). The number of additional interventional procedures to maintain TIPS patency was significantly greater in the bare-stent than in the PTFE-covered stent group (1.9 ± 1.2 vs. 0.6 ± 0.9; P = .007). The number of patients with clinical relapses was greater in the bare-stent group compared to the PTFE-covered stent group (13 vs. 5 episodes in 9 and 3 patients, respectively). In conclusion, PTFE-covered stents have a considerable advantage over bare stents for the TIPS treatment of BCS patients, with a lower dysfunction rate, a lower number of reinterventions, and fewer prosthesis requirements. PTFE-covered stents are preferable in patients with Budd-Chiari Syndrome. (HEPATOLOGY 2004;40:1197–1202.)
Budd-Chiari syndrome (BCS) is an uncommon liver disease defined as an obstruction to hepatic venous outflow at any level from the small hepatic veins to the junction of the inferior vena cava and the right atrium, excluding hepatic veno-occlusive disease.1 The obstruction of hepatic venous outflow leads progressively to sinusoidal congestion, centrilobular cell necrosis, and fibrosis, resulting in portal hypertension.2 Regardless of the etiology, consequences of portal hypertension, such as variceal bleeding, refractory ascites, hepatorenal syndrome, and spontaneous bacterial peritonitis, are very severe complications of BCS, with a potentially fatal outcome if not treated appropriately. Portal-systemic derivative procedures have been extensively used for patients unresponsive to conservative medical treatment, and surgical shunting has been the main strategy adopted over the last 20 years.3 However, no clear survival benefit has been demonstrated for surgical shunts, probably because of the great perioperative morbidity/mortality in such high-risk patients and the high incidence of shunt dysfunction (30%).1, 4–6 After the first reports of the successful use of transjugular intrahepatic portosystemic shunt (TIPS) for the treatment of Budd-Chiari syndrome in 1993,7, 8 surgical shunting has been progressively substituted in many centers with this less invasive approach.9–12 Indeed, it has been suggested that TIPS improves survival of BCS patients.12
Studies performed in patients with cirrhosis who are treated with TIPS have shown that TIPS dysfunction (thrombosis or stenosis of the shunt with the ensuing recurrence of portal hypertension complications) is a worrying complication that may occur in up to 60% of patients within 1 year of the TIPS procedure.13–16 Similar or even higher figures occur in BCS patients treated with TIPS.12 As happens when surgical shunts occlude, TIPS occlusion or dysfunction may have a deleterious effect on long-term survival.5, 17 In addition, TIPS dysfunction requires repeat invasive procedures to perform angioplasty or restenting, increasing patient discomfort, risks, and costs. It has recently been shown that the use of polytetrafluoroethylene (PTFE)-covered stents reduces the incidence of TIPS dysfunction in patients with cirrhosis who receive a TIPS for the treatment of variceal bleeding or refractory ascites.18, 19 However, whether these new PTFE-covered stents improve TIPS patency in patients with BCS has not been evaluated so far.
The aim of the present study was to analyze and compare the incidence of TIPS dysfunction and the outcome in 2 cohorts of patients with BCS who were treated with TIPS using bare or PTFE-covered stents.
The study included 25 TIPS procedures (16 bare stents and 9 PTFE-covered stents) in 21 BCS patients, who despite appropriate medical treatment developed refractory ascites and esophageal variceal bleeding and were considered to have BCS uncontrolled by medical therapy. Fifteen patients were treated at the Hepatic Hemodynamic Laboratory of the Barcelona Liver Unit from January 1993 to April 2003; they represent all patients with BCS treated with TIPS at that hospital. Twelve patients received uncovered stents (Wallstent; Boston Scientific, Natick, MA); after repeated episodes of TIPS dysfunction with clinical recurrence of portal hypertension (3 events for each patient), 4 were retreated by placing a PTFE-covered stent coaxially. Another 3 patients were treated with a PTFE-covered stent ab initio (Viatorr Endoprothesis; WL Gore, Flagstaff, AZ). Six additional BCS patients treated with TIPS from January 2001 to April 2003 at the Birmingham Liver Unit were included in the analysis. Four and 2 patients received bare and PTFE-covered stents, respectively. All patients had extensive thrombosis of at least 2 hepatic veins and the possibility of performing hepatic vein angioplasty or stenting was excluded. The etiology of BCS was systematically investigated in the patients as previously described.12, 20 Child-Pugh score, Rotterdam prognostic score index, and Clichy score index and its recently modified version were calculated at the time of TIPS insertion.5, 21, 22 TIPS insertion was performed as previously described.12 A continuous heparin infusion was initiated immediately after the portal vein was punctured, and an extrahepatic puncture site was ruled out. After 1 week, anticoagulation was shifted to warfarin and maintained for life. Once the etiological diagnosis was established, the specific treatment of the underlying disease, if present, was instituted.
Follow-up and TIPS Dysfunction.
Follow-up visits were scheduled monthly during the first 3 months, every 3 months for the first year, and every 6 months thereafter. Each visit included a clinical examination and laboratory tests. Visits at months 6, 12, 18, 24, and 30 included Doppler ultrasonography. Hemodynamic evaluations and portography were repeated every 6 months and whenever TIPS dysfunction was suspected clinically or at Doppler ultrasound, in accordance with the following criteria: no Doppler signal showing thrombosis of the stent, reduction of portal blood flow velocity below 28 cm/s−1, or reappearance of ascites.
TIPS dysfunction was always confirmed by hemodynamic study showing an increase of the portal pressure gradient above 12 mm Hg or by venography showing thrombosis of the stent, after which TIPS dysfunction was corrected by angioplasty or restenting.
End of follow-up was established at death, surgical shunt procedure, liver transplantation, or most recent evaluation, except for patients initially treated with bare stents who had repeated episodes of clinical recurrence and were subsequently treated with a PTFE-covered stent TIPS. These 4 patients were censored at insertion of the PTFE-covered stent.
The primary endpoint of the study was TIPS dysfunction. Secondary endpoints included the number of reinterventions (balloon dilatation or restenting) to accomplish TIPS patency, the number of stents required, and the number of clinical events during follow-up and survival.
Actuarial rates of shunt patency were estimated by the Kaplan-Meier method and compared in patients treated with bare and PTFE-covered prostheses using the log-rank test. Comparisons between groups were performed with the Student t test and the χ2 test when appropriate. Significance was established at a P value less than .05. All calculations were performed using the SPSS 10.0 statistical package (SPSS Inc., Chicago, IL).
Twenty-five TIPS procedures in 21 BCS patients (6 men and 15 women; mean age, 40 years; age range, 17-54) were included in the study. The hematological and thrombophilic diagnostic workup disclosed 10 patients with chronic myeloproliferative disorders, 6 patients with coagulopathy disorders, and 2 patients with paroxysmal nocturnal hemoglobinuria. Three patients remained idiopathic after an extensive hematological study. As shown in Table 1, no significant differences in the hepatic or renal function or the Budd-Chiari prognostic scores were observed at the time of TIPS treatment with bare or PTFE-covered stents. The indications for TIPS insertion were ascites, variceal bleeding, and liver insufficiency in the bare stent group (13/1/2, respectively) and PTFE-covered stent group (5/2/2, respectively). Mild clinical encephalopathy was present at the moment of TIPS insertion in 1 patient in the bare-stent group and in 3 patients in the PTFE-covered group. In addition, 2 patients in the bare-stent group exhibited mild encephalopathy during follow-up in conjunction with TIPS dysfunction. Mild encephalopathy remained in only 1 patient in the PTFE-covered group despite TIPS patency.
Table 1. Hepatic and Renal Function and Budd-Chiari Prognostic Scores at Time of TIPS Treatment With Bare or PTFE-Covered Stents
No significant differences were found between any of the parameters analyzed.
2.4 ± 1.0
3.6 ± 3.2
Prothrombin index (%)
56 ± 16
52 ± 21
Serum albumin (g/L)
32 ± 3
32 ± 9
Serum creatinine (mg/dL)
1.1 ± 0.3
0.9 ± 0.3
9 ± 2
9 ± 2
5.8 ± 1.0
6.0 ± 1.1
6.2 ± 1.9
6.9 ± 1.4
1.3 ± 0.6
1.4 ± 0.8
Eleven patients in the bare-stent group (69%) had a puncture from the right hepatic vein stump to the right portal branch vein, 3 patients (19%) had a direct puncture from the intrahepatic portion of the inferior vena cava to the right portal vein and one to the left portal vein, and the remaining patient had a puncture from the right hepatic vein stump to the left portal vein. In the PTFE group, 6 patients (66%) had a puncture from the right hepatic vein stump to the right portal vein, and the other 3 patients (33%) had a puncture from the inferior vena cava to the right portal vein. Therefore, no substantial technical differences between the groups were observed. In this series of patients, no complications related to the procedure were observed. The median follow-up period for the bare and PTFE-covered TIPS group were 21.2 months (range, 3.9-124.8) and 19.1 months (range, 7.7-31.3), respectively (not significant).
Fourteen of 16 patients (87%) who received bare stents developed TIPS dysfunction during the follow-up compared to 3 of the 9 patients (30%) who received PTFE-covered stents (P = .005). As shown in Fig. 1, the actuarial rates of primary patency in the bare-stent group were 19% at 1 year and 9% at 2 years compared with 67% at 1 and 2 years in the PTFE-covered stent group (P = .02; log-rank test). This beneficial effect of PTFE stents was also evident when those PTFE patients initially treated with bare stents were not included twice in the analysis and censored at the time of PTFE stenting. In this subgroup, the actuarial rates of primary patency at 1 year were 18% in the bare-stent group and 60% in the PTFE-covered stent group (P = .1).
In the bare-stent group, first TIPS dysfunction was clinically apparent in 6 cases (recurrence of ascites in 5 and of variceal bleeding in 1) and detected by follow-up hemodynamic evaluations in the other 8 cases (Fig. 2). TIPS dysfunction was corrected with angioplasty in 4 cases and restenting with bare stents in 9 cases (Fig. 2). Due to complete obliteration with thrombosis of the bare stent in the remaining patient, a new stent placement was technically unfeasible, and the patient was censored from the study at 6.1 months. He was still alive, with minimal ascites, after 2 years of follow-up. A second episode of TIPS dysfunction appeared during follow-up in 12 of these 13 patients (92%), manifesting as recurrent ascites in 4 and asymptomatic in the remaining 8. Restenting was performed in 6 cases and angioplasty alone in 5. In the remaining patient—despite confirmation of stent thrombosis—the demonstration by ultrasonography of spontaneous development of large collaterals associated with the absence of clinical symptoms led to an expectant attitude. Seven of 16 patients had more than 2 episodes of TIPS dysfunction. Only 2 patients (12.5%) treated with bare stents did not develop TIPS dysfunction after a mean follow-up of 22.7 months (13 and 33 months).
TIPS dysfunction occurred in 3 cases with PTFE-covered stents and was manifested by recurrent ascites. Dysfunction was corrected by insertion of a new coaxial stent. Of the 3 cases, 2 developed a second episode of clinical TIPS dysfunction 2 months later, requiring restenting in one and angioplasty in the other; the other patient persists as asymptomatic and without TIPS dysfunction after a total follow-up of 31 months. Six of 9 patients (66%) were free of TIPS dysfunction after a mean follow-up of 20.8 ± 7.0 months.
The number of interventional procedures to maintain TIPS patency was significantly greater in the bare stent than in the PTFE group (1.9 ± 1.2 vs. 0.6 ± 0.9; P = .007). The number of stents used during follow-up was also significantly greater in the bare group than in the PTFE group (Table 2).
Table 2. Mean Number of Stents, Invasive Interventions, Rehospitalizations, and Days in Hospital in Each Treatment Group
PTFE-Covered Stent Group
NOTE. Data are expressed as mean ± SD.
Stents during follow-up
1.6 ± 1.7
0.4 ± 0.7
Number of angioplasties
0.7 ± 0.7
0.1 ± 0.3
Number of restenting procedures
1.2 ± 1.2
0.4 ± 0.7
4.1 ± 1.3
2 ± 1.1
Overall days in hospital
13 ± 12
4 ± 5
Clinical relapses were more frequent in the bare-stent group compared to the PTFE-covered stent group (13 episodes in 9 patients vs. 5 episodes in 3 patients, respectively), although this difference did not reach statistical significance.
Two patients, both with bare stents, finally underwent orthotopic liver transplantation because of progressive deterioration of liver function. No patients died during the follow-up period.
Bare stents used for the treatment of BCS are extremely efficient in the short term. However, TIPS dysfunction is a common problem associated with reappearance of clinical events and requiring multiple restenting and angioplasty sessions to restore TIPS patency. Indeed, the reported incidence of shunt dysfunction ranges between 40% and 75% in 3 TIPS series for BCS after an average follow-up of 26 months.12, 23, 24 These figures are greater than those reported for patients with cirrhosis who receive a TIPS for the treatment of portal hypertension (an average of 50% at 1 year14) and are probably influenced by the multiple coexisting underlying prothrombotic disorders present in BCS patients.25 In patients with cirrhosis, the use of PTFE-covered stents has recently been shown to maintain long-term shunt patency, reducing the rate of dysfunction, the number of reintervention requirements, and the number of clinical events compared to patients who receive bare stents.19, 26–30
This study compared 2 cohorts of severe Budd-Chiari patients treated with TIPS using bare or PTFE-covered stents, with a detailed and prospective long-term follow-up. The results of the present study clearly show that the use of PTFE-covered stents is associated with a marked and significant reduction in the incidence of TIPS dysfunction compared to that seen with the use of bare stents. Correspondingly, a significantly lower number of reinterventions, stents, and balloon angioplasty procedures were required in the PTFE group. This was observed without significant differences in the length of follow-up between both groups.
A trend toward a reduced number of clinically relevant TIPS dysfunction episodes was observed in the PTFE-covered group. It is possible that our close follow-up and early management may have prevented the appearance of additional clinically relevant dysfunction events in the bare-stent group, as has been suggested by other investigators,18, 31 but these events would have increased both the cost of the treatment and patient discomfort. Therefore, we estimate that the reduced reintervention rates and prosthesis requirements, the lower number of clinical recurrence, hospital admissions, and overall days in the hospital largely counterbalance the higher costs of the PTFE stents. However, this should be confirmed by additional specific studies.
Four patients that were censored from the bare-stent group, after repeated TIPS dysfunction with recurrence of severe symptoms, were treated with PTFE stents and included as new cases in the PTFE group. However, similar results were observed when these 4 patients were excluded from the PTFE group (primary patency rate, 18% vs. 60%). These results strongly support the contention that PTFE-covered stents provide better results for TIPS not only for cirrhosis but also for BCS. In addition, they emphasize the feasibility and efficacy of PTFE-covered stents used not only for de novo indications but also for the correction of TIPS dysfunction in patients with severe clinical recurrence.26, 30, 32
Importantly, it should be noted that no Budd-Chiari patient in this series who was treated with a PTFE-covered stent developed liver infarcts. Liver infarcts have been reported after PTFE stents in patients with liver cirrhosis and are due, in most cases, to the obstruction of the hepatic vein by the covered stent. The lack of liver infarcts in our patients is not unexpected. Indeed, hepatic veins are already occluded in Budd-Chiari patients, and in most of our cases the liver puncture to the portal vein went through the liver parenchyma (departing from a hepatic vein stump and sometimes from the intrahepatic portion of the vena cava) and not through a hepatic vein.
In conclusion, this study showed a substantial benefit of PTFE-covered TIPS over bare stents for the treatment of clinically symptomatic BCS patients, with lower dysfunction rates and fewer reinterventions and prostheses used, and suggests that PTFE-covered stents should be used preferentially to treat these patients.
The authors thank Ms. M. A. Baringo, L. Rocabert, and R. Saez for their expert technical assistance, and M. Montaño for editorial support.