Polytetrafluoroethylene Covered Stent: Results of an Italian Multicenter Study

TIPS with Expanded Covered Stent Interventional Radiology Original Research Claudio Vignali 1 Irene Bargellini 1 Maurizio Grosso 2 Giovanni Passalacqua 3 Franco Maglione 4 Fulvio Pedrazzini 2 Pietro Filauri 3 Raffaella Niola 4 Roberto Cioni 1 Pasquale Petruzzi 1 Vignali C, Bargellini I, Grosso M, et al. Received July 23, 2004; accepted after revision October 19, 2004. 1 Division of Diagnostic and Interventional Radiology, Department of Oncology, Transplants, and Advanced Technologies in Medicine, University of Pisa, Via Roma 67, Pisa 56127, Italy. Address correspondence to C. Vignali (cvignali@med.unipi.it). 2 Department of Radiology, S. Croce e Carle Hospital, Via Coppino 26, Cuneo 12100, Italy. 3 Department of Radiology, Santissimi Filippo e Nicola Hospital, Località Tre Conche, Avezzano 67051, Italy. 4 Department of Radiology, Cardarelli Hospital, Via A. Cardarelli no. 9, Napoli 80131, Italy. AJR 2005; 185:472 480 0361 803X/05/1852 472 American Roentgen Ray Society TIPS with Expanded Polytetrafluoroethylene Covered Stent: Results of an Italian Multicenter Study OBJECTIVE. Our objective is to describe the results of a multicenter prospective trial on the safety and efficacy of transjugular intrahepatic portosystemic shunts (TIPS) using the Viatorr stent-graft. SUBJECTS AND METHODS. From 2001 to 2003, 114 patients (75 men and 39 women; mean age, 59.3 years) with portal hypertension underwent TIPS with the Viatorr stent-graft. Indications for treatment were variceal bleeding (n = 49, 43.0%), refractory ascites (n = 52, 45.6%), hypertensive gastropathy (n = 10, 8.8%), Budd-Chiari syndrome (n = 1, 0.9%), and hepatorenal syndrome (n = 2, 1.7%). Eight patients (7.0%) had Child-Pugh class A cirrhosis; 60 (52.6%), Child-Pugh class B; and 46 (40.4%), Child-Pugh class C. Patients were monitored by color Doppler sonography and phlebography. RESULTS. The procedure was successful in 113 (99.1%) of 114 patients; in one patient, creation of the track was not feasible. The mean portosystemic pressure gradient decreased from 21.8 to 8.7 mm Hg. Three minor immediate complications (2.6%) occurred (two cases of self-limiting hemoperitoneum and one extrahepatic portal puncture requiring covered stenting). At a mean follow-up of 11.9 months, the overall mortality rate was 31.0% (35/113), with a 30-day mortality rate of 8.8% (10/113). Mortality was significantly higher in patients in Child-Pugh class C with refractory ascites and with postprocedural encephalopathy. Cumulative primary patency rates were 91.9%, 79.9%, and 75.9% at 6, 12, and 24 months follow-up, respectively. Restenosis occurred in 15 patients (13.3%) within the stent (n = 8, 53.3%) or at the ends of the portal (n = 1, 6.7%) or hepatic (n = 6, 40%) veins and was solved by percutaneous transluminal angioplasty (n = 11), stenting (n = 3), or parallel TIPS (n = 1). The secondary patency rate was 98.2%. Postprocedural encephalopathy occurred in 27 patients (23.9%). CONCLUSION. The Viatorr stent-graft is safe and effective in TIPS creation, with high primary patency rates. Covering the entire track up to the inferior vena cava can increase patency. n the last decade, the effectiveness I of transjugular intrahepatic portosystemic shunts (TIPS) in managing complications of portal hypertension has been shown by several clinical studies [1 4]. Initial trials focused on the use of bare metal stents, whose major drawback was the high rate of shunt obstruction [5 7], requiring close, invasive, and costly surveillance and frequent revisions. In several randomized trials comparing TIPS with other treatment options, a significant reduction of rebleeding rates [2, 8, 9] and improvement of ascites [10] were shown; however, no survival benefits seemed to be associated with TIPS [11, 12]. Therefore, TIPS has been indicated in patients awaiting liver transplantation [13] or as a rescue procedure when other treatments have failed [12]. In the past few years, covered stent-grafts have been introduced as a valid alternative to bare stents in TIPS, and animal studies [14, 15] and initial clinical trials have shown high primary patency rates and significant survival benefits [16, 17]. The graft material should be biocompatible, nonthrombogenic, and impermeable to bile and should provide a substrate for the endothelial lining [16]. Initial experiences were with custom-made devices. However, in recent years, a new covered stent (Viatorr GORE-TEX vascular graft, W. L. Gore and Associates) has become commercially avail- 472 AJR:185, August 2005

TIPS with Expanded Covered Stent able in Europe; initial larger clinical trials have been published, reporting high technical success rates and a low incidence of restenosis at midterm follow-up [18 21]. The Viatorr stent-graft is a covered stent specifically designed for TIPS [17]. It is an expanded polytetrafluoroethylene (eptfe)- covered nitinol stent-graft consisting of two parts: a distal 2-cm-long uncovered portion that lies in the portal vein and an inner covered portion (4 8 cm long) for the track and the hepatic side of the TIPS. The eptfe is a trilaminar structure with slight porosity and impermeable to liquid bile, and the metal skeleton of the stent-graft consists of a 0.2-mm nitinol wire with a zigzag configuration. The stent-graft is available in 8-, 10-, and 12-mm diameters. This article presents the results of an Italian multicenter trial evaluating the safety, efficacy, and long-term performance of the Viatorr stent-graft in the creation and revision of TIPS for relieving portal hypertension and its complications. Subjects and Methods This prospective multicenter study was performed at four Italian hospitals. The study protocol was approved by the institutional ethics committees of all centers involved. Informed consent was obtained from all patients. Patients From February 2001 to December 2003, 114 consecutive patients (75 men and 39 women; age range, 18 83 years; mean, 59.3 ± 12.1 [SD] years) experiencing complications from portal hypertension were enrolled. Primary indications for TIPS included variceal bleeding (n = 49, 43%), refractory ascites (n =52, 45.6%), hypertensive gastropathy (n = 10, 8.8%), Budd-Chiari syndrome (n = 1, 0.9%), and hepatorenal syndrome (n = 2, 1.7%), all of which were refractory to or intolerant of conventional endoscopic or medical therapies. In patients in whom refractory ascites was the indication for TIPS, examination of ascitic fluid revealed no evidence of spontaneous bacterial peritonitis or other infection. No patient had signs of systemic infection. The causes of hepatic disease were hepatitis (n = 58), alcohol abuse (n = 36), both hepatitis and alcohol abuse (n = 14), Budd-Chiari syndrome (n = 2), a cryptogenic cause (n = 1), metastasis (n = 1), and portal vein thrombosis (n = 2) (Fig. 1). A B C D Fig. 1 18-year-old man with portal vein thrombosis and variceal bleeding (patient 3, Table 2). Patient had previously undergone splenectomy for lymphoma. A C, Angiographic images in anteroposterior projection demonstrate transjugular intrahepatic portosystemic shunt procedure. Left portal vein was punctured under sonographic and fluoroscopic guidance, and injection of contrast material showed presence of portal cavernoma with opacification of right portal branch (A, arrow). Track was created between right hepatic vein and right portal branch (B), and Viatorr stent-graft (diameter, 10 mm; length, 80 mm; W. L. Gore and Associates) was successfully deployed (C), with reduction of portosystemic pressure gradient from 19 to 7 mm Hg. D E, Phlebographic control images obtained 6 months after procedure show severe stenosis of portal vein, distal to stent-graft margin (D). Patient was asymptomatic, but high portosystemic pressure gradient was observed (20 mm Hg). Memotherm stent (diameter, 12 mm; length, 80 mm; Bard Peripheral Vascular) was successfully deployed (E). Stent-graft was patent and patient asymptomatic at 34 months follow-up. E AJR:185, August 2005 473

Seven patients had undergone previous orthotopic liver transplantation. Eight patients had Child-Pugh class A cirrhosis; 60, Child-Pugh class B; and 46, Child-Pugh class C. Preprocedural Workup Before the TIPS procedure, all patients underwent complete clinical and laboratory assessments. Sonography and color-coded duplex sonography followed by contrast-enhanced MDCT were performed on all patients to visualize the liver parenchyma and exclude focal liver lesions, to assess the presence and extent of complications of portal hypertension, to assess the vascular anatomy, and to select the most proper hepatic vein for the procedure. Inclusion and Exclusion Criteria All incoming patients who had portal hypertension complications refractory to or intolerant of medical or endoscopic therapies and who had adequate vascular access at the level of the hepatic veins were considered for inclusion in the study. Exclusion criteria were chronic severe encephalopathy, severe right cardiac failure, and multifocal hepatocellular carcinoma. TIPS Procedure, Stent-Graft Implantation, and Portosystemic Pressure Gradient (PPG) Measurement The Viatorr stent-graft and TIPS procedure have been described elsewhere [17]. The procedure was performed in the angiographic suite while the patients were deeply sedated. The present series included only elective TIPS creation. The right jugular vein was punctured under sonographic guidance. A standard 10-French, 40-cm-long Rösch-Uchida TIPS set (Cook Europe) was used for TIPS creation. In most patients, the right hepatic vein and the right portal branch were used to create the track, under sonographic and fluoroscopic guidance. After dilatation of the track with high-pressure balloon catheters, a venogram was obtained with a calibrating catheter to assess the track length, select the proper stent-graft, and facilitate precise deployment. All stents were 10 mm in diameter; they varied in length. Once the stent-graft had been deployed, the stent was dilated with high-pressure balloons of sizes equivalent to the nominal diameters of the grafts. Shunt venography was performed to monitor shunt patency and diameter and the persistence of varices. When required, variceal coil embolization was performed at the end of the procedure. When venography showed incomplete track coverage, an adjunctive covered stent was deployed. The PPG was recorded before the procedure, after track creation, and at TIPS completion. Follow-Up After the procedure, all patients underwent strict clinical follow-up. Stent patency was monitored by color-coded duplex sonography at discharge; at 1, 3, 6, and 12 months after TIPS; and annually thereafter and by portography at 6 months follow-up or in cases of suspected shunt malfunction. Portography was performed either by a right jugular or by a common femoral venous access. PPG was always recorded. In some instances, MDCT was also required for a better depiction of the liver and vascular anatomy and of the stent-graft position and patency. In cases of complications, further percutaneous treatments were required. Medication Protocol All patients received broad-spectrum antibiotic medication as perioperative prophylaxis, starting 2 days before the procedure and continuing for 1 week. No anticoagulation protocol was followed in this series, either during the procedure or postprocedurally. Definition of Success Immediate technical success was defined as creation of the TIPS, with an associated 50% reduction in the preprocedural PPG or a PPG of less than 12 mm Hg [22]. In-stent stenosis was defined as a 50% reduction in the stent-graft lumen or by a PPG increase of more than 12 mm Hg. In-graft and above-the-graft stenoses were recorded separately. Primary and secondary patency rates and survival rates were calculated on follow-up. Immediate and long-term TIPS-related complications, such as encephalopathy and recurrence of bleeding or ascites, were recorded. Statistical Analysis Descriptive statistics (proportions, means, SDs, and medians) were calculated for patient characteristics and numeric continuous data. The chi-square, Pearson, and Fisher exact tests were used for categoric data, whereas continuous data were analyzed by bivariate fit, one-way analysis of variance, and Student s t test. Cumulative patency and survival rates were calculated with the Kaplan-Meier method, and groups were compared using the log-rank test (chi-square) or the Wilcoxon rank sum test. Stepwise multiple logistic regression analyses were performed to assess factors influencing shunt malfunction and mortality. A p value of less than 0.05 was considered statistically significant. Statistical analysis was performed using a statistical software program (JMP; SAS Institute Inc.). Results Procedural Results Shunt insertion was successful in 113 (99.1%) of 114 patients. In one patient, the track could not be created because the vessels were severely tortuous and hypotrophic. Mean stent-graft length was 8 ± 1 mm; the following lengths were used: 4 mm in two patients, 6 mm in nine patients, 7 mm in 17 patients, 8 mm in 38 patients, 9 mm in 40 patients, and 10 mm in seven patients. In 108 (95.6%) of 113 patients, the stent-graft was placed accurately with no evidence of immediate complications. In the remaining five patients, covering the entire track required the use of an adjunctive covered stent, represented by a Viatorr stent-graft in three patients and by a Jostent (Abbott Vascular Devices) in two patients. The Jostent was used to cover a remaining track length of 2 3 mm because this size is not available with the Viatorr stent-graft. The adjunctive stent was required in two patients to cover a long track and in the remaining patients because of misjudgment of the track length, despite the use of a calibrated catheter. At completion of the procedure, variceal coil embolization was considered necessary and was successfully performed on 11 patients. Mean preprocedural PPG was 21.8 ± 5.6 mm Hg (range, 13 34 mm Hg). After the procedure, mean PPG was 8.7 ± 3.7 (range, 2 17 mm Hg), with a mean reduction of 60.3% ± 12.5% (range, 36 88%). TIPS-Related Complications Three minor immediate complications occurred: Two cases of self-limiting hemoperitoneum and one extrahepatic portal puncture, sealed with an adjunctive stent-graft. One patient, who had Child-Pugh class C hepatitis C virus cirrhosis, ascites, hypertensive gastropathy, and hepatocellular carcinoma, experienced severe hematemesis and melena 10 hr after TIPS; the patient underwent portography, which showed no evidence of stent-graft occlusion or other complications, despite the finding of high PPG (16 mm Hg). Multiple angioplasties were performed within the stent and at the portal and hepatic veins margins; final PPG was 4 mm Hg. However, the patient died from multiorgan failure 2 days later. No contrast medium related complications occurred. No puncture-related major complications or infectious complications were observed. 474 AJR:185, August 2005

TIPS with Expanded Covered Stent Shunt Patency and Additional Procedures Mean follow-up was 11.9 ± 10.2 months (range, 0 38 months). According to the Kaplan-Meier analysis, cumulative primary patency rates were 91.9%, 79.9%, and 75.9% at follow-up intervals of 6, 12, and 24 months, respectively (Fig. 2). TIPS restenosis occurred in 15 (13.3%) of 113 patients either within the stent (n =8, 53.3%) or at the extremity of the portal (n =1, 6.7%) (Fig. 1) or hepatic vein (n =6, 40%). The mean grade of stenosis was 72% ± 19.2% (range, 30 100%). Six patients were asymptomatic, five (4.4%) showed persistent or recurrent ascites, and the remaining four (3.5%) underwent further evaluation because of recurrent variceal bleeding (Table 1). Among patients with clinical recurrence, postprocedural PPG was greater than 12 mm Hg in only one patient, with variceal recurrent bleeding. Portography performed on these patients showed suboptimal coverage of the track at the hepatic vein extremity in all cases of hepatic vein restenosis and in seven of eight cases of in-stent restenosis. In one patient (patient 13, Table 2), with less than 30% early restenosis, we observed incomplete dilatation of the stent-graft with kinking in the midportion of the stent. Patients were treated by percutaneous transluminal angioplasty (n = 11) or by a noncovered stent (n = 3; two Jostent and one Memotherm stent [Bard Peripheral Vascular]) (Table 1). One patient (patient 15, Table 2) required repeated percutaneous transluminal angioplasty within 1 month, resulting in technical and clinical success. Fig. 2 Kaplan-Meier analysis for primary patency. Cumulative primary patency rate at 24 months follow-up was 75.9%. Dotted lines indicate confidence intervals. In one patient with persistent ascites (patient 13, Table 2), duplex sonography and MDCT showed graft displacement and complete thrombosis; stent-graft catheterization was not feasible by either femoral or jugular access, and creation of a new shunt was required (Fig. 3). One patient (patient 8, Table 2), with variceal rebleeding requiring TIPS revision, died from hepatic failure 1 week after repeated percutaneous transluminal angioplasty. One patient (patient 1, Table 2) showed persistent ascites after the secondary procedure and underwent liver transplantation 2 months later. Therefore, the secondary success rate was 98.2% (111/113). Shunt malfunction was significantly more frequent in male patients than in female patients (p = 0.05), younger patients (p = 0.03), or patients in whom shorter stents had been used (p = 0.002). In fact, mean stent length in the group of patients with shunt dysfunction was 7.2 mm, compared with 8.2 mm in patients without complications. On the stepwise multiple regression analysis, stent length and the age and sex of the patient were independent risk factors for shunt malfunction. In particular, stent length represented the most important risk factor (p < 0.006; odds ratio, 49.5; 95% confidence interval, 0.2 1.1). Ten patients underwent liver transplantation 1 18 months after TIPS (mean, 7 ± 6 months); in all these patients, inspection of the stent-graft did not reveal in-stent obstruction or restenosis. Mortality and Other Complications The 30-day mortality rate was 8.8% (10/113). Causes of early mortality were TABLE 1: Patients Clinical and Demographic Characteristics Variable Value Age (yr) 59.3 ± 12.1 Male patient (%) 65.8 Cause of liver disease (%) Hepatitis 50.9 Alcohol abuse 31.6 Hepatitis and alcohol abuse 12.3 Other a 5.2 Ascites (%) 45.6 Laboratory tests Albumin (g/dl) 2.4 ± 0.8 Bilirubin (mg/dl) 3.3 ± 5.2 Creatinine (mg/dl) 1.6 ± 1.4 International normalized ratio 1.5 ± 0.4 Child-Pugh score 9.4 ± 1.8 Child-Pugh class (%) A 7.0 B 52.6 C 40.4 Portosystemic pressure gradient (mm Hg) Before transjugular intrahepatic 21.8 ± 5.6 portosystemic shunt After transjugular intrahepatic 8.7 ± 3.7 portosystemic shunt Note Data are mean ± SD, unless otherwise indicated. a Budd-Chiari syndrome, cryptogenic cause, portal vein thrombosis, and liver metastasis. acute hepatic failure (n = 5) and disseminated intravascular coagulation (n =5), with no clear cases of procedure-related deaths. The overall mortality rate was 31.0% (35/113). Cumulative survival rates were 77.8%, 68.6%, and 59.3% at 6, 12, and 24 months, respectively (Fig. 4). Mortality showed a statistically significant relationship with Child-Pugh class C (p = 0.0003, Wilcoxon rank sum test), the presence of refractory ascites (p =0.03, Wilcoxon rank sum test), and the presence of postprocedural encephalopathy (p =0.05, Wilcoxon rank sum test) (Figs. 5 7). No statistically significant relationship was observed between mortality and patient age or sex, cause of hepatopathy, pre- or postprocedural PPG, and shunt malfunction during follow-up. Encephalopathy was observed in 27 (23.9%) of 113 patients on follow-up. However, it had been preprocedurally diagnosed AJR:185, August 2005 475

in nine of 27 patients, with no worsening after TIPS. In only five patients was it severe (grade 3). All patients were responsive to medical treatment alone and did not require in-stent reduction. The postprocedural PPG ranged from 2 to 8 mm Hg in 10 of 27 patients, whereas in the remaining patients it was higher than 8 mm Hg and in five patients it was higher than 12 mm Hg. No statistically significant difference in the immediate postprocedural PPG was observed between patients with and patients without postprocedural encephalopathy. Discussion The role of TIPS in the treatment of complications related to portal hypertension has been widely assessed [1 4]. The most frequently reported complication after TIPS is in-stent restenosis, occurring in up to 50% of patients at 12-months follow-up [5 7]. Three models of shunt stenosis or occlusion have been reported [16]. Early stenosis may be caused by bile duct transection and biliary fistulas [5, 23], whereas later occlusions are caused by an inflammatory healing response to the stent and proliferation of fibroconnective tissue [24]. Finally, intimal hyperplasia can develop in the hepatic veins above the stent, contributing to shunt dysfunction [6, 25]. Ideally, stent-grafts should address all three types of stenosis, and the graft material should provide specific features to reduce bile leakage and fibroblast proliferation and to favor endothelial lining [17]. Several graft materials have been proposed. Nishimine et al. [14] and Haskal et al. [15] reported the benefits of lining the stents with eptfe in porcine models, with an increased patency from 8% to 50% at 1 month s follow-up [14] and no bile staining in TIPS stent-grafts [15]. Other materials showed less promising results [26 29]. The Viatorr stent-graft uses an eptfe lining that is biocompatible, microporous, nonthrombogenic, and relatively impermeable to bile and tissue and provides a substrate for endothelial lining [16, 17]. Initial studies reported high patency rates with this new, commercially available device [18 21]. Ideally, the increased patency would allow reduction of invasive follow-up and therefore reduce costs. Moreover, the longer durability of the stent-graft seems to improve survival [18]. In the present series, incomplete track coverage represented the main cause of shunt dysfunction [25], having been observed in all six cases of hepatic vein stenosis above the graft and in seven of the eight cases (7%) of in-graft restenosis. Portography performed on these seven patients revealed, at the level of the hepatic vein, the presence of incomplete track coverage that might cause displacement and kinking of the stent-graft, inducing thrombosis. This finding is in accordance with recently published data supporting the need for optimal and precise placement of the stent-graft up to the inferior vena cava with complete track coverage [16, 17, 30]. No complications related to complete coverage of the hepatic veins, such as partial Budd-Chiari syndrome or hepatic infarction, were observed in our series, although some cases have been reported in the literature [31]. Of a series of 20 patients, Otal et al. [31] reported two cases of partial Budd-Chiari syndrome with no clinical effects. It seems that hepatic venous occlusion is unlikely to be clinically significant because of the development of venous collateral circulation [32]. Puncturing the hepatic vein as close as possible to the inferior vena cava could reduce the risk of hepatic vein occlusion [31] and the incidence of hepatic vein stenosis and stent displacement [17]. In addition, in our series a statistically significant relationship was observed between the incidence of shunt dysfunction and the use of short stents. However, it is also extremely important not to use excessively long stents, particularly in patients suitable for liver transplantation, because coverage of the portal vein close to the hilum or coverage of the inferior vena cava might interfere with vascular clamping during transplantation [33]. Shunt dysfunction can be managed safely by percutaneous procedures, such as balloon angioplasty and adjunctive stenting [20, 21, 25], with high technical success rates. In our series, only one patient could not be re-treated by angioplasty or stenting, and a parallel TIPS was required. Up to 100% secondary patency rates have been reported, with low periprocedural complications [16, 17, 19, 21]. TABLE 2: Shunt Malfunction During Follow-Up Patient No. Follow-Up (mo) Stenosis Site Stenosis Grade (%) Symptoms Treatment Secondary Success 1 2 Hepatic vein 70 Ascites PTA Persistent ascites 2 22 Hepatic vein 40 None PTA Yes 3 6 Portal vein 90 None Stent Yes 4 12 Hepatic vein 70 None PTA Yes 5 12 Stent 100 Bleeding PTA Yes 6 12 Hepatic vein 70 None PTA Yes 7 6 Stent 60 Bleeding PTA Yes 8 10 Stent 60 None PTA Yes 9 12 Stent 80 Bleeding PTA Death 10 12 Stent 75 Bleeding PTA Yes 11 5 Hepatic vein 80 Ascites Stent Yes 12 14 days Stent 30 Ascites PTA Yes 13 12 Stent 100 Ascites Repeated TIPS Yes 14 2 Hepatic vein 75 None Stent Yes 15 1 Stent 80 Ascites PTA Repeated PTA Note PTA = percutaneous transluminal angioplasty, TIPS = transjugular intrahepatic portosystemic shunt. 476 AJR:185, August 2005

TIPS with Expanded Covered Stent Fig. 4 Kaplan-Meier analysis for survival. Mean expected survival was 21.9 months, with 59.3% cumulative overall survival at 24 months follow-up. A B C D Fig. 3 42-year-old man with Budd-Chiari syndrome and variceal bleeding (patient 13, Table 2). A D, Angiograms show success of transjugular intrahepatic portosystemic shunt (TIPS) procedure performed with Viatorr stent-graft (diameter, 10 mm; length, 80 mm; W. L. Gore and Associates) (A). One year after treatment, patient underwent MDCT for ascites; complete stent-graft occlusion was observed (B, multiplanar reformation in coronal plane) associated with incomplete track coverage at level of hepatic vein (C), volume rendering in coronal-oblique plane). Catheterization of stent-graft was not feasible, requiring creation of a new track; angiography after placement of TIPS showed correct deployment of Viatorr stent-graft (diameter, 10 mm; length, 70 mm) followed by coil embolization of a large varix (D). Fig. 5 Kaplan-Meier analysis for survival grouped by Child-Pugh class. Survival rate was 100% in patients with Child-Pugh class A and was significantly higher for class A than for class B or C (log-rank test, p = 0.0004). In the present study, clinical recurrence on follow-up was low, with only four cases (3.5%) of variceal recurrent bleeding and 5 cases (4.4%) of persistent ascites, all of which were associated with shunt dysfunction. These results compare favorably with previous studies, which reported 3 9% clinical relapse [19 21, 25]. In the prospective setting of the present study, close surveillance by duplex sonography and venography was required. However, sonography and duplex sonography are AJR:185, August 2005 477

Fig. 6 Kaplan-Meier analysis for survival grouped by postprocedural encephalopathy. Survival rate was significantly higher in patients without postprocedural encephalopathy (log-rank test, p = 0.008). highly accurate and, particularly when associated with proper clinical and laboratory evaluation, might be sufficient to evaluate stent-graft patency and the presence of complications [34]. CT might represent a valuable diagnostic tool as well, particularly to visualize stent-graft displacement, although the use of iodinated contrast medium and irradiation and the higher costs may limit its role to specific cases. Venography still represents the only imaging technique able to measure PPG, yet it is highly invasive, costly, and resource demanding. The increased primary patency and reduced shunt dysfunction associated with the use of covered stents might eventually reduce the need for invasive imaging and favor noninvasive surveillance, with obvious benefits to patient management and costs. No anticoagulation medication was used in our series. In fact, we could find no reported data supporting benefits from oral anticoagulation or antiplatelet medication alone on TIPS patency rates [35, 36]. The high cumulative primary patency rate (75.9% at 2 years) reported in our series might be a further confirmation that no anticoagulation protocol is needed, particularly in patients with liver dysfunction and a low platelet count, in whom the risks of anticoagulation seem to be higher than the supposed risks of graft stenosis. Our data confirm the safety of TIPS and its high immediate technical success rate [21]. No major immediate complications were observed in our study, and the only case of severe melena, although of unclear cause, was not related to the procedure or to stent-graft dysfunction. However, important complications have been reported during and immediately after TIPS, such as severe hemorrhage and graft infections in cases of bacteremia [37]. We believe that accurate patient selection is desirable whenever possible, to reduce the incidence of complications. Accurate procedural planning is possible using duplex sonography and MDCT with adequate postprocessing, to allow precise definition of the vascular and liver anatomy. High immediate technical success rates can be achieved after adequate training of the operators. Puncture of the portal vein requires specific skills and knowledge of the anatomy and of the technique; when possible, sonographic guidance should be used to avoid puncture of important liver structures and multiple punctures of the extrahepatic track of the portal vein. Moreover, a calibrated catheter always should be used to evaluate track length: The length of the external curvature of the track (which is usually the longest one) represents the minimum stent-graft length required. A stent-graft length of 8 cm is usually recommended, with shorter devices deployed when the track is created between the left hepatic and left portal veins. The early and overall mortality rates in our series are in accordance with published data, and the cumulative survival rate at 2 years follow-up 59.3% is lower than in other reports [18, 21]. However, in our series a higher Fig. 7 Kaplan-Meier analysis for survival grouped by main indication for placement of transjugular intrahepatic portosystemic shunt (variceal bleeding or refractory ascites). Survival rate was significantly higher in patients who underwent shunt placement for variceal bleeding than in those with refractory ascites (log-rank test, p =0.02). rate (36%) of Child-Pugh class C patients was treated than in previous studies. Rossi et al. [21] reported an 82.7% survival rate at 1 year s follow-up, with only 18% of patients in Child-Pugh class C, whereas Angermayr et al. [18], 29% of whose patients were in Child-Pugh class C, reported a cumulative survival rate of 76% at 2 years follow-up. In fact, Child-Pugh class represents an important predictor of survival [18]. As expected, most deaths in our series occurred in patients with severe cirrhosis (Child-Pugh class C) [38]. Recent reports have also focused on the usefulness of the MELD (model for end-stage liver disease) score to predict survival in patients after TIPS [39]. Moreover, as already pointed out in previous reports, survival was lower in patients who underwent TIPS for the treatment of refractory ascites than in patients whose indication for TIPS was variceal bleeding [18, 39]. In fact, ascites is known to be a poor prognostic factor in patients with cirrhosis, and its clinical management appears to be problematic. In a recent study, Ferral et al. reported a poor survival rate in patients with a MELD score of less than 17 and refractory ascites and suggested that TIPS creation should be avoided in the treatment of refractory ascites in patients with advanced liver failure [39]. In our series, another independent risk factor for survival was postprocedural encephalopathy, which has been reported to be a major drawback of TIPS, with reported incidences of 25 52% using bare stents [4] and 31 44% 478 AJR:185, August 2005

TIPS with Expanded Covered Stent using the Viatorr stent-graft [20, 21]. No significant difference in the incidence of encephalopathy was found between bare and covered stents [19, 20]. In our series, encephalopathy occurred in 23.9% of patients, although it was severe in only five patients; all cases could be controlled by medical therapy. However, strictured stents have been proposed to reduce TIPS diameter and thus encephalopathy [40]. In our series, patients were treated exclusively by 10-mm-diameter stent-grafts, in an attempt to reduce the incidence of encephalopathy yet maintain acceptable patency. The goal of TIPS is to reduce PPG to less than 12 mm Hg, thus reducing the risk of recurrent bleeding but increasing the incidence of encephalopathy. However, our series could not show a relationship between clinical relapse and high postprocedural PPG or between encephalopathy and low PPG, with five cases of encephalopathy occurring in patients with a PPG of more than 12 mm Hg. To reduce the incidence of shunt-induced hepatic encephalopathy and liver failure, Rossle et al. [22] proposed that the goal of TIPS be to reduce PPG by 25 50%, rather than to reduce it to the widely used threshold value of 12 mm Hg. In defining our immediate technical success rate, we were agreeing with this finding. The main limitations of our study were the relatively short follow-up period and the high number of patients lost during it. In fact, 12 patients (10.6%) did not comply with the required continuous follow-up and were lost to follow-up a few months after treatment. Randomized prospective trials comparing covered and bare stents in TIPS are now needed to weigh the benefits of stent-grafts, in terms of patency and long-term survival, against their higher costs. These trials are under way. Were the benefits of covered stents to be confirmed, further studies comparing TIPS and medical and surgical therapies would be needed to establish the proper management protocol for patients with portal hypertension and its complications. To our knowledge, this is the largest published series on the use of the Viatorr stentgraft for TIPS creation. This off-the-shelf, commercially available TIPS stent-graft is easy to handle and safe in selected patients. When adequately placed, the stent-graft does not interfere with eventual liver transplantation. In a follow-up of up to 38 months, the high primary patency rate was confirmed. In most patients, restenosis can be treated safely and effectively by further percutaneous procedures. Bridging the complete track to the inferior vena cava could result in higher patency rates, reducing the need for reinterventions and therefore making TIPS a more economic treatment for portal venous hypertension and its associated complications, despite the higher cost of the graft [16, 17]. References 1. Richter GM, Noeldge G, Palmaz JC, et al. Transjugular intrahepatic portocaval stent shunt: preliminary clinical results. Radiology 1990; 174:1027 1030 2. Rossle M, Deibert P, Haag K, et al. Randomised trial of transjugular-intrahepatic-portosystemic shunt versus endoscopy plus propranolol for prevention of variceal rebleeding. Lancet 1997; 349:1043 1049 3. Ochs A, Rössle M, Haag K, et al. The transjugular intrahepatic portosystemic stent shunt procedure for refractory ascites. N Engl J Med 1995; 332:1192 1197 4. ter Borg PC, Hollemans M, Van Buuren HR, et al. Transjugular intrahepatic portosystemic shunts: long-term patency and clinical results in a patient cohort observed for 3 9 years. Radiology 2004; 23:537 545 5. Saxon RR, Ross PL, Mendel-Hartvig J, et al. Transjugular intrahepatic portosystemic shunt patency and the importance of stenosis location in the development of recurrent symptoms. Radiology 1998; 207:683 693 6. Nazarian GK, Ferral H, Castaneda-Zuniga WR, et al. Development of stenoses in transjugular intrahepatic portosystemic shunts. Radiology 1994; 192:231 234 7. Domagk D, Patch D, Dick R, et al. Transjugular intrahepatic portosystemic shunt in the treatment of portal hypertension using Memotherm stents: a prospective multicenter study. Cardiovasc Intervent Radiol 2002; 25:506 512 8. Sanyal AJ, Freedman AM, Luketic VA, et al. Transjugular intrahepatic portosystemic shunts compared with endoscopic sclerotherapy for the prevention of recurrent variceal hemorrhage: a randomized, controlled trial. Ann Intern Med 1997; 126:849 857 9. Jalan R, Forrest EH, Stanley AJ, et al. A randomized trial comparing transjugular intrahepatic portosystemic stent-shunt with variceal band ligation in the prevention of rebleeding from esophageal varices. Hepatology 1997; 26:1115 1122 10. Rosemurgy AS, Zervos EE, Clark WC, et al. TIPS versus peritoneovenous shunt in the treatment of medically intractable ascites: a prospective randomized trial. Ann Surg 2004; 239:883 889 11. Rossle M, Ochs A, Gulberg V, et al. A comparison of paracentesis and transjugular intrahepatic portosystemic shunting in patients with ascites. N Engl JMed 2000; 342:1701 1707 12. Papatheodoridis GV, Goulis J, Leandro G, Patch D, Burroughs AK. Transjugular intrahepatic portosystemic shunt compared with endoscopic treatment for prevention of variceal rebleeding: a meta-analysis. Hepatology 1999; 30:612 622 13. Ring EJ, Lake JR, Roberts JP, et al. Using transjugular intrahepatic portosystemic shunts to control variceal bleeding before liver transplantation. Ann Intern Med 1992; 116:304 309 14. Nishimine K, Saxon RR, Kichikawa K, et al. Improved transjugular intrahepatic portosystemic shunt patency with PTFE-covered stent-grafts: experimental results in swine. Radiology 1995; 196:341 347 15. Haskal ZJ, Davis A, McAllister A, Furth EE. PTFE-encapsulated endovascular stent-graft for transjugular intrahepatic portosystemic shunts: experimental evaluation. Radiology 1997; 205:682 688 16. Haskal ZJ. Improved patency of transjugular intrahepatic portosystemic shunts in humans: creation and revision with PTFE stent-grafts. Radiology 1999; 213:759 766 17. Cejna M, Peck-Radosavljevic M, Thurnher SA, Hittmair K, Schoder M, Lammer J. Creation of transjugular intrahepatic portosystemic shunts with stent-grafts: initial experiences with a polytetrafluoroethylene-covered nitinol endoprosthesis. Radiology 2001; 221:437 446 18. Angermayr B, Cejna M, Koenig F, et al. Survival in patients undergoing transjugular intrahepatic portosystemic shunt: eptfe-covered stentgrafts versus bare stents. Hepatology 2003; 38:1043 1050 19. Bureau C, Garcia-Pagan JC, Otal P, et al. Improved clinical outcome using polytetrafluoroethylenecoated stents for TIPS: results of a randomized study. Gastroenterology 2004; 126:469 475 20. Hausegger KA, Karnel F, Georgieva B, et al. Transjugular intrahepatic portosystemic shunt creation with the Viatorr expanded polytetrafluoroethylene-covered stent-graft. J Vasc Intervent Radiol 2004; 15:239 248 21. Rossi P, Salvatori FM, Fanelli F, et al. Polytetrafluoroethylene-covered nitinol stent-graft for transjugular intrahepatic portosystemic shunt creation: 3-year experience. Radiology 2004; 231:820 830 22. Rossle M, Siegerstetter V, Olschewski M, Ochs A, Berger E, Haag K. How much reduction in portal pressure is necessary to prevent variceal rebleeding? A longitudinal study in 225 patients with transjugular intrahepatic portosystemic shunts. Am J Gastroenterol 2001; 96:3379 3383 23. Teng GJ, Bettmann MA, Hoopes PJ, et al. Transjugular intrahepatic portosystemic shunt: effect of AJR:185, August 2005 479

bile leak on smooth muscle cell proliferation. Radiology 1998; 208:799 805 24. Sanyal AJ, Contos MJ, Yager D, Zhu YN, Willey A, Graham MF. Development of pseudointima and stenosis after transjugular intrahepatic portosystemic shunts: characterisation of cell phenotype and function. Hepatology 1998; 28:22 32 25. Angeloni S, Merli M, Salvatori FM, et al. Polytetrafluoroethylene-covered stent grafts for TIPS procedure: 1-year patency and clinical results. Am J Gastroenterol 2004; 99:280 285 26. Haskal ZJ, Brennecke LH. Transjugular intrahepatic portosystemic shunts formed with polyethylene terephthalate-covered stents: experimental evaluation in pigs. Radiology 1999; 213:853 859 27. Otal P, Rousseau H, Vinel J, Ducoin H, Hassissene S, Joffre F. High occlusion rate in experimental transjugular intrahepatic portosystemic shunt created with a Dacron-covered nitinol stent. JVasc Intervent Radiol 1999; 10:183 188 28. Ferral H, Alcantara-Peraza A, Kimura Y, Castaneda-Zuniga WR. Creation of transjugular intrahepatic portosystemic shunts with use of the Cragg Endopro System I. J Vasc Intervent Radiol 1998; 9:283 287 29. Haskal ZJ, Weintraub JL, Susman J. Recurrent TIPS thrombosis after polyethylene stent-graft use and salvage with polytetrafluoroethylene stent-grafts. J Vasc Intervent Radiol 2002; 13:1255 1259 30. Clark TW, Agarwal R, Haskal ZJ, Stavropoulos SW. The effect of initial shunt outflow position on patency of transjugular intrahepatic portosystemic shunts. J Vasc Intervent Radiol 2004; 15:147 152 31. Otal P, Smayra T, Bureau C, et al. Preliminary results of a new expanded-polytetrafluoroethylene-covered stent-graft for transjugular intrahepatic portosystemic shunt procedures. AJR 2002; 178:141 147 32. Cho KJ, Geisinger KR, Shields JJ, Forrest ME. Collateral channels and histopathology in hepatic vein occlusion. AJR 1982; 139:703 709 33. Maleux G, Pirenne J, Vaninbroukx J, Aerts R, Nevens F. Are TIPS stent-grafts a contraindication for future liver transplantation? Cardiovasc Intervent Radiol 2004; 27:140 142 34. Zizka J, Elias P, Krajina A, et al. Value of Doppler sonography in revealing transjugular intrahepatic portosystemic shunt malfunction: a 5-year experience in 216 patients. AJR 2000; 175:141 148 35. Theilmann L, Sauer P, Roeren T, et al. Acetylsalicylic acid in the prevention of early stenosis and occlusion of transjugular intrahepatic portal-systemic stent shunts: a controlled study. Hepatology 1994; 20:592 597 36. Siegerstetter V, Huber M, Ochs A, Blum HE, Rossle M. Platelet aggregation and platelet-derived growth factor inhibition for prevention of insufficiency of the transjugular intrahepatic portosystemic shunt: a randomized study comparing trapidil plus ticlopidine with heparin treatment. Hepatology 1999; 29:33 38 37. DeSimone JA, Beavis KG, Eschelman DJ, Henning KJ. Sustained bacteremia associated with transjugular intrahepatic portosystemic shunt (TIPS). Clin Infect Dis 2000; 30:384 386 38. Angermayr B, Cejna M, Karnel F, et al. Child-Pugh versus MELD score in predicting survival in patients undergoing transjugular intrahepatic portosystemic shunt. Gut 2003; 52:879 885 39. Ferral H, Gamboa P, Postoak DW, et al. Survival after elective transjugular intrahepatic portosystemic shunt creation: prediction with model for end-stage liver disease score. Radiology 2004; 231:231 236 40. Cox MW, Soltes GD, Lin PH, Bush RL, Lumsden AB. Reversal of transjugular intrahepatic portosystemic shunt (TIPS)-induced hepatic encephalopathy using a strictured self-expanding covered stent. Cardiovasc Intervent Radiol 2003; 26:539 542 480 AJR:185, August 2005