Chapter 21: Transjugular Intrahepatic Shunts for Variceal Bleeding: Results of Clinical Studies Variceal hemorrhage is a well-recognized and frequently lethal complication of portal hypertension (PHT). In nearly 50% of patients with cirrhosis, gastroesophageal varices are present at the time of diagnosis.1 These varices grow at a rate of 7% per year, and the 1-year rate of sentinel bleeding is 12%.2,3 When bleeding occurs, patients have an 80% chance of rebleeding within 1 year and a mortality rate of bleeding episodes approaching 33%.4 Since its inception the transjugular intrahepatic shunt (TIPS) procedure has evolved in portal access techniques, shunt technology, image guidance, and shunt surveillance. Because of these refinements, TIPS has largely supplanted most surgical shunting procedures as a means of long-term portal decompression. This chapter summarizes our current understanding of the role of TIPS for variceal bleeding. Over the past 20 years, TIPS has increasingly become a vital tool in reducing PHT for preventing variceal rebleeding. Rosch et al initially described this conduit as a durable means for treating variceal bleeding while avoiding open surgery.5 More than 10 randomized prospective trials have compared TIPS with endoscopic therapy (ET) in the prevention of variceal rebleeding.6 The initial trials conducted in the mid to late 1990s with bare-metal stents (BMS) demonstrated that patients treated with TIPS rebled at a lower frequency than those treated with ET (18.9% vs. 46.6%, respectively).7 This benefit was juxtaposed by a higher rate of hepatic encephalopathy in TIPS patients (34% vs. 19%, respectively).7 Most of the trials also indicated no difference in overall mortality rate between the two procedures, yet there were consistently fewer deaths caused by rebleeding in the TIPS group (2.4% vs. 7.7%).7–11 A recent meta-analysis that collected data from 12 randomized controlled trials (RCTs) and four previous meta-analyses confirmed these results (decreased rebleeding rates, increased encephalopathy, no change in overall mortality rate with TIPS) and reported no difference in the number of hospital days between the two treatment arms.12 Escorsell et al compared TIPS with pharmacologic therapy (propanolol + isosorbide-5-mononitrate) and found similar results as endoscopy trials. The TIPS arm had lower rebleeding probability (13% vs. 49%), higher rates of encephalopathy (38% vs. 14%), and no difference in survival probability.13 Each of these trials was limited by high rates of crossover from the endoscopic or medical arms to rescue therapy with TIPS, which may have masked differences in mortality rate as patients who would have died from recurrent variceal bleeding in the endoscopic arm were shunted with TIPS. These initial RCTs also raised recognition that the BMS used for TIPS were prone to dysfunction, which led to the recurrence of PHT and subsequent bleeding ( These recurrences developed in 25% to 50% of patients after 6 to 12 months.14–17 Dysfunction rates accumulated over time and were reported as high as 70% after 2 years in several trials.12,18 The etiology of dysfunction was hypothesized to be multifactorial. Potential causes included thrombosis from stent migration or biliary fistula ( Fig. 21.1 A 51-year-old man with alcoholic cirrhosis and previous transjugular intrahepatic portosystemic shunt (TIPS) for recurrent variceal bleeding. (a) The TIPS was created with a bare stent 4 months earlier, and the stent has retracted into the hepatic parenchymal tract. Catheter advanced to the area of stenosis (arrows) where TIPS joins hepatic vein. (b) Additional stent placed across shunt stenosis and dilated with balloon catheter. The need for better stents for TIPS with longer patency rates drove the innovation of polytetrafluoroethylene (PTFE)-covered stent grafts. In 1995, Nishimine et al reported the first use of these stent grafts in pig studies with subsequent improvement in patency rates.22 Saxon et al reported initial clinical results in 1997 of the use of stent grafts for revision of TIPS. They reported an initial mean primary patency of 50 days compared with a mean of 229 days after graft placement.15 Haskal achieved similar results in 1999 and showed a mean patency duration of 19 months in a series of 13 patients.19 These initial PTFE grafts appropriately blunted physiologic responses such as myofibroblast invasion and extracellular collagen deposition while also preventing bile leaks.19 In 1999, W. L. Gore & Associates developed the first commercially available stent graft using expanded PTFE (ePTFE) marketed as VIATORR ( In 2003, Angermayr et al performed the first large retrospective study comparing BMS with stent graft patency rates and survival data. Three-month, 1-year, and 2-year survival rates were reported as 93%, 88%, and 76% for the ePTFE group and 83%, 73%, and 62% for the BMS group (n = 89 for stent graft, n = 419 BMS).24 This represented the first case-control study that not only supported the increased technical patency of TIPS stent grafts but also indicated improved survival with stent grafts. However, studies by Hausegger et al and Charon et al did not support these claims of improved mortality rates. They reported 1-year survival rates of 72% and 65%, respectively, for stent grafts with no statistical difference between groups.25,26 Stent grafts were expected to have higher rates of encephalopathy, but most retrospective studies showed no significant difference between the two types of stents.25,26,27 The reported incidence of recurrent bleeding with stent grafts was 3.7% to 7%, compared with rebleeding rates of 18% to 20% in the BMS.7,25,26 Fig. 21.3 The VIATORR (W. L. Gore, Flagstaff, Arizona) device is a polytetrafluoroethylene (PTFE) conduit supported by a nitinol stent skeleton. The portal venous end of the device is uncovered to permit unimpeded portal blood flow through the stent interstices. Although there are only a minimal number of RCTs comparing stent grafts with BMS, the results of these trials are concordant with earlier retrospective studies ( It is clear that ePTFE-covered stents are superior to their bare-metal precursors. The convincing data led the American Association for the Study of Liver Diseases to update their TIPS practice guidelines in 2009 to recommend the use of stent grafts for TIPS.33 It is well known that the standard of care for primary prevention of variceal bleeding remains β-blockers because surgical shunts or TIPS pose significant incremental risks.33 No studies have been completed to assess the efficacy of TIPS for primary prevention. Acute hemorrhage from ruptured varices represents a clinical emergency. If not adequately treated, the mortality rate can approach 80%.34 Patients must be immediately stabilized and treated with pharmacologic therapy and ET. Spontaneous hemostasis only occurs in 50% of patients.35 If bleeding persists despite initial therapy, emergent TIPS is recommended for portal decompression. The effectiveness of TIPS in controlling refractory bleeding is tempered by a high mortality rate. The urgency of the procedure along with the worse clinical status of patient (model for end-stage liver disease [MELD] score, Acute Physiology and Chronic Health Evaluation [APACHE] II score, and Child-Pugh class) account for this higher mortality. Specifically, Child-Pugh class C; elevated hepatic venous pressure gradient (HPVG, greater than 20 mm Hg); hemodynamic instability; and other comorbidities such as heart failure, sepsis, and multiorgan failure contribute to poor outcomes.36,37 Vangeli et al compiled data from 15 studies in which TIPS was used to control refractory bleeding. Immediate control was achieved in 94% with rebleeding seen in 12%, but the mortality rate was high at 6 weeks, approaching 36% to 50%.38 A large case series published by Gazzera et al further supported this high first month mortality rate seen in emergent TIPS patients. These investigators correlated mortality to Child class C, elevated creatinine and prolonged prothrombin time and not to TIPS technique.34
Introduction
Prevention of Variceal Rebleeding: Early Clinical Studies
Fig. 21.1).
Fig. 21.2), parenchymal stenosis from fibrotic healing, and intimal hyperplasia from chronic injury.14,15,19–21 These technical problems mandated more intense clinical and imaging follow-up, with frequent reinterventions and higher costs.21 Although TIPS was proven to be superior to ET in the prevention of rebleeding, the drawbacks of greater encephalopathy, stent complications, and lack of demonstrable survival benefit led to the recommendation that ET remain the standard of care as first-line therapy for patients with variceal bleeding.
The Era of Stent Grafts
Fig. 21.3). Before its approval by the United States Food and Drug Administration (FDA), several European series confirmed the ability of the device to maintain patency for mean follow-up times of 282 to 387 days.21,23
Table 21.1). The clinical study performed for PMA (premarket approval) by the FDA of the VIATORR device for TIPS was a multicenter RCT with 253 subjects who underwent de novo TIPS procedure with either VIATORR or a commercially available bare-metal endoprosthesis (Boston Scientific Wallstent). Variceal bleeding was the primary indication in 35% to 36% of these patients. Primary patency success was defined as portosystemic pressure gradient (PSG) of less than 12 mm Hg and percent diameter stenosis less than 50%. At 6 months, primary patency rates were 60% and 22.7% for the VIATORR and control group, respectively. There was no significant difference between time to return of symptoms, but the VIATORR group had significantly fewer shunts with stenosis and reduced need for shunt revisions.28 Bureau et al performed a smaller RCT with 80 patients that also demonstrated better patency, less reintervention, decreased encephalopathy, and similar long-term survival with TIPS stent grafts.29,30 Primary patency of covered and BMS groups were 86% and 47%, respectively, at 1 year, and 80% and 19% at 2 years. A more recent meta-analysis by Yang et al also confirmed previous reports and indicated a trend toward better survival with PTFE stent grafts.31
Current Practice Guidelines
Primary Prevention of Variceal Bleeding
Acute Bleeding
Thoracic Key
Fastest Thoracic Insight Engine
