DOI: 10.1148/rg.241035028
(Radiographics. 2004;24:21-36.)
© RSNA, 2004
TIPS-related Hepatic Encephalopathy: Management Options with Novel Endovascular Techniques1
David C. Madoff, MD,
Michael J. Wallace, MD,
Kamran Ahrar, MD and
Richard R. Saxon, MD
1 From the Division of Diagnostic Imaging, Section of Vascular and Interventional Radiology, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Unit 325, Houston, TX 77030-4009 (D.C.M., M.J.W., K.A.); and North County Radiology, Vista, Calif (R.R.S.). Presented as an education exhibit at the 2002 RSNA scientific assembly. Received February 6, 2003; revision requested April 23 and received June 5; accepted June 9. Address correspondence to D.C.M. (e-mail: dmadoff@di.mdacc.tmc.edu).
 |
Abstract
|
|---|
Hepatic encephalopathy is a common complication that develops after creation of a transjugular intrahepatic portosystemic shunt (TIPS). Although most patients respond well to conservative medical therapy (ie, protein-restricted diet, nonabsorbable disaccharides, nonabsorbable antibiotics), a small percentage of patients (3%–7%) do not benefit from these methods and require more invasive therapeutic approaches. One option is emergent liver transplantation, but the majority of patients are not suitable candidates. Recently, various percutaneous techniques have been described that alter the hemodynamics through the shunt by occluding it with coils or balloons or by reducing its diameter by inserting constrained stents or stent-grafts. Other techniques have been used for patients with TIPS-related hepatic encephalopathy in whom spontaneous splenorenal shunts are present. In many patients with refractory hepatic encephalopathy, these percutaneous techniques have produced symptomatic improvement, with either a complete resolution or a substantial reduction in hepatic encephalopathy symptoms that can be controlled with medical therapy. Unfortunately, despite all attempts, some patients remain incapacitated and ultimately die. Further research is necessary to improve our understanding of TIPS-related hepatic encephalopathy so that newer, less invasive and safer procedures can be developed to treat this difficult clinical problem.
© RSNA, 2004
Index Terms: Brain, diseases, 10.458 • Hypertension, portal, 957.711 • Liver, interventional procedures, 761.126, 957.458, 957.1268 • Shunts, portosystemic, 761.126, 957.458, 957.1268
|
LEARNING OBJECTIVES FOR TEST 2
|
|---|
After reading this article and taking the test, the reader will be able to:
- Discuss the proposed pathophysiology that leads to hepatic encephalopathy after TIPS.
- Identify the management options for patients who develop hepatic encephalopathy after TIPS.
- Describe endovascular techniques that can be performed for patients whose hepatic encephalopathy is refractory to medical management.
|
Introduction
|
|---|
Hepatic encephalopathy, a complication that commonly develops in patients after creation of a transjugular intrahepatic portosystemic shunt (TIPS), is characterized by confusion, disorientation, obtundation, abnormal sleep patterns, and overall alterations in quality of life (1). New or worsened hepatic encephalopathy after TIPS has been reported to occur in 5%–35% of patients, but conservative medical management usually is sufficient to reverse the problem (2). Hepatic encephalopathy that is refractory to these forms of treatment develops in 3%–7% of patients (3) for who further intervention, including shunt occlusion or reduction and potentially emergent liver transplantation, may be required.
Various percutaneous techniques used to treat this difficult clinical dilemma alter the hemodynamics through the shunt by occluding or reducing its diameter. Complete shunt occlusion, accomplished with coils or balloons, may have abrupt, life-threatening hemodynamic consequences (4–6). Other techniques that attempt to reduce flow by creating turbulence within the shunt rely on the placement of constrained, uncovered stents within the TIPS lumen (3,7–9). Results of the latter approach are unpredictable, and several days are often required for the portosystemic gradient (PSG) to increase and stabilize. More recently, endografts constrained to a predetermined diameter have shown promise as an effective means of shunt reduction that results in an immediate and measurable hemodynamic response (10).
In this article, we review the proposed pathophysiology that leads to hepatic encephalopathy after TIPS and the medical management options for patients with this condition. In addition, we present illustrations and radiographic images of endovascular techniques that can be performed for patients in whom hepatic encephalopathy is refractory to conservative therapy. The advantages, disadvantages, and potential consequences of these novel approaches are discussed. This article discusses the use of endovascular devices for the management of refractory hepatic encephalopathy that are, at present, considered off-label by the Food and Drug Administration.
|
Post-TIPS Hepatic Encephalopathy
|
|---|
Although the exact pathophysiology of hepatic encephalopathy is complex and poorly understood, the belief is that central nervous system alterations arise when intestinally derived compounds requiring hepatic detoxification bypass the liver and remain in the systemic circulation (Fig 1) (1,3,11,12). Which toxin or toxins lead, directly or indirectly, alone or in combination, to the development of hepatic encephalopathy is a topic of much debate (13). Among many theories, the most widely accepted one is that nitrogenous compounds (ammonia in particular) enter the systemic circulation because of decreased hepatic function or by portosystemic shunting. Once within the brain tissue, the compounds cause alterations in neurotransmission that lead to disturbances in consciousness and behavior. Other substances proposed to cause hepatic encephalopathy include 
-aminobutyric acid–benzodiazepines, true neurotransmitters (eg, glutamate, norepinephrine, dopamine), false neurotransmitters–plasma amino acid imbalance, short- and medium-chain fatty acids, phenols, mercaptans, serotonin-tryptophan, manganese, and endogenous opioids. These chemicals may interact with ammonia to result in additional neurologic disturbances (14).
View larger version (73K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Diagram illustrates the proposed complex feedback mechanisms that can lead to hepatic encephalopathy. BCAA/AAA = branched chain-aromatic amino acids, BZD = benzodiazepines, DA = dopamine, GABA = -aminobutyric acid, GLU = glutamic acid, 5HT = serotonin, SCFA = short chain fatty acids.
|
|
The pathogenesis of post-TIPS hepatic encephalopathy is related to a combination of increased bioavailability of gut-derived toxins and portal hypoperfusion (13,15). Two phenomena are believed to contribute to diminished portal venous flow: diversion of portal venous flow away from the liver (Fig 2) or, secondarily, "sinusoidal steal" that results in further cellular compromise in the face of preexisting hepatocellular dysfunction (11).
View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Changes in hemodynamics after TIPS creation. (a) Pre-TIPS venogram demonstrates antegrade blood flow through the portal venous system and retrograde blood flow through esophageal varices (arrow). (b) Post-TIPS venogram shows that the majority of blood flows through the TIPS, bypassing the hepatic parenchyma. Note the decrease in blood flow through the esophageal varices (arrow).
|
|
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Changes in hemodynamics after TIPS creation. (a) Pre-TIPS venogram demonstrates antegrade blood flow through the portal venous system and retrograde blood flow through esophageal varices (arrow). (b) Post-TIPS venogram shows that the majority of blood flows through the TIPS, bypassing the hepatic parenchyma. Note the decrease in blood flow through the esophageal varices (arrow).
|
|
|
Management of Hepatic Encephalopathy
|
|---|
Medical Therapy
Episodes of hepatic encephalopathy that occur within weeks after TIPS are often induced by specific precipitating factors (eg, dietary protein indiscretion, gastrointestinal bleeding, sepsis, dehydration, hypokalemia, hypoxia, constipation, use of sedatives or psychoactive medications), particularly in patients with ascites and in patients who take diuretics after TIPS (1,12). Post-TIPS diuresis with rapid weight loss may result in electrolyte imbalance and intravascular volume depletion that may precipitate hepatic encephalopathy. Acute deterioration of liver function in cirrhosis that leads to hepatic encephalopathy may also result from superimposed alcoholic hepatitis or an acute circulatory disturbance such as portal vein thrombosis (14).
Hepatic encephalopathy that develops after TIPS may forecast the development of chronic recurrent encephalopathy. In these circumstances, precipitating factors considered reversible might be successfully treated with protein-restricted diets, nonabsorbable disaccharides (eg, lactulose), or antibiotics (eg, neomycin, metronidazole). Drugs that affect neurotransmission may have a therapeutic role in selected patients.
The medical management of post-TIPS hepatic encephalopathy is similar to the approaches used to treat hepatic encephalopathy in patients who did not undergo TIPS. In patients without other precipitating factors, evaluation of the TIPS and a search for large splenorenal or gastrorenal shunts should be performed.
Dietary Management.—
The classic cornerstone of therapy for acute hepatic encephalopathy has been restriction of dietary protein, with subsequent incremental increases to assess clinical tolerance (14). The dietary management of hepatic encephalopathy includes increasing dietary fiber to promote catharsis and restricting protein intake to no less than 0.5 g of protein per kilogram of body weight per day. This regimen may worsen the overall nutritional status of these already malnourished patients, adding to the potential morbidity associated with liver transplantation should that option become available. Importantly, a positive nitrogen balance will have beneficial effects on hepatic encephalopathy by promoting liver regeneration and by improving the capacity of the muscles to detoxify ammonia. The decision to use severe dietary protein restriction must be carefully considered, with the risks of worsening hepatic encephalopathy weighed against malnutrition (16). Branched chain-amino acids (BCAA) may provide a better-tolerated source of protein (17). However, the high costs and limited therapeutic value of BCAA should be reserved for patients with hepatic encephalopathy refractory to maximal medical therapy.
Zinc, a co-factor of urea cycle enzymes, may be deficient in patients with cirrhosis, especially if malnutrition is present. As zinc deficiency has been reported to precipitate hepatic encephalopathy, use of zinc supplements has been advocated (18).
Reduction in Nitrogenous Load from the Gut.—
Nitrogenous load in the gut may be reduced by means of bowel cleansing, administration of nonabsorbable disaccharides, and use of nonabsorbable antibiotics.
Bowel cleansing has been used because toxins responsible for the development of hepatic encephalopathy are derived from the gut. Hepatic encephalopathy itself may result in slow transit time, and colonic cleansing with various laxatives and enemas or irrigation with a 5-L isotonic mannitol solution has been used to reduce intraluminal ammonia, intraluminal bacterial content, and blood ammonia (14,19–21).
Nonabsorbable disaccharides, such as lactulose, have been proposed to work by causing catharsis and acidification of the colon (22), although the precise mechanisms of their action are not known. Lactulose is not metabolized by intestinal disaccharidases, so that it enters the colon. Acidification of the colon occurs as colonic bacteria ferment the nonabsorbable disaccharide into acetic acid and lactic acid, which leads to passage of ammonia from the bloodstream into the colonic lumen. Catharsis assists in the removal of neurotoxins from the gut.
Lactulose may be administered orally (ingested or via nasogastric tube) or rectally, depending on the patient’s mental status and gastrointestinal integrity. The dose of orally administered lactulose—generally 15–45 mL every 8–12 hours—is titrated to achieve two to three semiformed bowel movements per day (14). When rectal administration is used, 300 mL of lactulose in 1 L of water is administered while the patient remains in the Trendelenburg position for 1 hour to increase the likelihood that the lactulose reaches the ascending colon (23). Side effects include volume depletion from diarrhea, electrolyte imbalances, and renal insufficiency, all of which may worsen hepatic encephalopathy.
Nonabsorbable antibiotics are used as second-line therapy for patients who do not respond to lactulose. Neomycin has traditionally been used to reduce the population of urease-containing bacteria in the colonic lumen during acute episodes of hepatic encephalopathy (24,25). However, although neomycin is poorly absorbed, 1%–3% reaches the systemic circulation and may lead to neurotoxicity, ototoxicity, and renal failure. In addition, neomycin affects the small bowel mucosa, impairing the activity of glutaminase in the intestinal villi. Intestinal malabsorption may result in spruelike diarrhea. For these reasons, use of neomycin should be avoided in long-term management of patients with hepatic encephalopathy. Metronidazole may be used as an alternative agent, but its long-term use is also limited by the potential development of irreversible neuropathy (14,26,27).
In some cases, the treatment may be gradually tapered off during follow-up as hepatic encephalopathy improves. Gradual improvement may be, in part, related to progressive shunt stenosis. Hepatic encephalopathy that occurs in the presence of shunt stenosis and recurrent variceal hemorrhage poses a clinical dilemma. If encephalopathy is disabling, an alternative endovascular approach to treat variceal hemorrhage that would prevent the worsening of hepatic encephalopathy could include variceal embolization without shunt revision. Alternatively, endoscopic treatment or liver transplantation can be used, depending on the severity of the liver dysfunction (1).
Drugs That Affect Neurotransmission.—
Flumazenil and bromocriptine may have a therapeutic role in selected patients with hepatic encephalopathy, based on the hypothesis that transmission of -aminobutyric acid (GABA) is enhanced in development of hepatic encephalopathy (14,28). Agents that bind to GABAA-receptors exert direct neuroinhibitory effects on the brain, leading to symptoms of hepatic encephalopathy (29). Because the antagonistic effects of flumazenil were found to have modest effectiveness in one large clinical trial (30) and because flumazenil has been reported occasionally to cause seizures, formal recommendations for using this drug cannot be made (14).
Dopaminergic alterations have also been postulated to cause hepatic encephalopathy, especially after recent observations (31) that manganese accumulated in the basal ganglia of patients with cirrhosis. Extrapyramidal signs may frequently be seen in patients with liver disease (32). These signs are reported to improve following bromocriptine administration (33), although bromocriptine may cause prolactin levels to rise.
Endovascular Therapy
A small but important subset of patients who do not respond to medical management will require a more invasive approach (1,34). Endovascular treatment for these patients is directed toward decreasing the amount of portal venous blood diverted away from the liver and occluding (either permanently or reversibly) or reducing (either with uncovered or covered constrained stents) the existing TIPS.
Shunt Occlusion.—
Permanent shunt occlusion has been used to treat post-TIPS hepatic encephalopathy and fulminant liver failure (4,35). In 1984, Potts et al (36) first described insertion of a detachable balloon to occlude a surgically created splenorenal shunt. Following shunt occlusion, the authors reported a decrease in cardiac output and an increase in vascular resistance and hepatic perfusion, along with measurably improved hepatocyte function. Based on this successful precedent, interventional radiologists have used embolic agents (stainless steel coils and detachable balloons) for permanent TIPS occlusion (Fig 3).
Despite the successful clinical use of shunt occlusion, several subsequent reports have described life-threatening sequelae of this approach. Aside from the immediate increase in portal venous pressure and risk of recurrent variceal hemorrhage, other severe hemodynamic consequences have been encountered. In 1994, Paz-Fumagalli et al (4) reported a case in which intentional TIPS occlusion with coils resulted in death. Although no postmortem evaluation was performed, death was believed to have resulted from sudden, severe hemodynamic alterations that caused decreased cardiac output, hypotension, and metabolic acidosis. The insertion of permanent foreign objects (coils and detachable balloons) into a TIPS to reverse shunt-related complications could pose a problem should variceal hemorrhage recur. In these cases, straightforward recanalization of the shunt might not be feasible and creation of a new TIPS might be required.
Intentional reversible TIPS occlusion also has been described as a way to treat refractory hepatic encephalopathy. Kerlan et al (6) and Haskal et al (5) reported five and two cases, respectively, of successful reversible thrombosis achieved by placing completely occlusive latex balloons (Meditech/Boston Scientific, Watertown, Mass) within the midportion of the TIPS for up to 48 hours (Fig 4). The technique was used to cause thrombosis of the transparenchymal portion of the shunt below the balloon. The advantage of this technique is its reversibility, should ascites or variceal bleeding recur. If recanalization is performed, a smaller diameter shunt may be used to reduce the risk of recurrent hepatic encephalopathy. As with any technique that abruptly occludes the portosystemic shunt, this approach also restores the substantial risk of recurrent variceal hemorrhage and may produce life-threatening hemodynamic changes (4–6,36). The technique also poses the theoretic risk of thrombus propagation, either proximally or distally, into the portal or hepatic vein. Another potential complication of this approach is balloon migration into the right side of the heart or balloon rupture. To avoid balloon migration, a guide wire may be left within the balloon catheter lumen, or a long (8-F) sheath may be placed within the shunt through which the balloon catheter is inserted (5). Periodic radiographic imaging during the balloon inflation process is prudent to ensure that these problems are identified and treated.
View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Temporary TIPS occlusion. (a) Drawing shows an occlusion balloon within the TIPS, with development of thrombus below (portal side). (b) Thrombus remains within the intraparenchymal tract following removal of the occlusion balloon. (c) If necessary, recanalization can be performed to reestablish flow through the TIPS.
|
|
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Temporary TIPS occlusion. (a) Drawing shows an occlusion balloon within the TIPS, with development of thrombus below (portal side). (b) Thrombus remains within the intraparenchymal tract following removal of the occlusion balloon. (c) If necessary, recanalization can be performed to reestablish flow through the TIPS.
|
|
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Temporary TIPS occlusion. (a) Drawing shows an occlusion balloon within the TIPS, with development of thrombus below (portal side). (b) Thrombus remains within the intraparenchymal tract following removal of the occlusion balloon. (c) If necessary, recanalization can be performed to reestablish flow through the TIPS.
|
|
Shunt Reduction.—
To overcome the problems associated with shunt occlusion, several techniques have been developed that diminish flow by creating turbulence within the shunt lumen. These techniques include use of constrained stents, adjunct embolization, and constrained stent-grafts.
Haskal and Middlebrook (7) reported on use of a constrained stent that resulted in significant clinical improvement in a patient with refractory hepatic encephalopathy. They employed a Wallstent (Boston Scientific, Natick, Mass), which they constrained with a 3–0 silk suture to create an hourglass-shaped stent with a constrained or reduced lumen diameter of 5 mm. The authors attributed the reduction in blood flow through the stent to the increased friction and turbulence created by the interposed stent mesh. Their technique (Fig 5) involved the insertion of a 12-F sheath (Meditech/Boston Scientific) into the right internal jugular vein. A 40-cm-long 9-F sheath was then backloaded onto the Wallstent delivery catheter, and the stent was partially deployed until its leading half was exposed. A 3–0 silk suture was woven through the meshwork of the stent and tied, creating a waist in the stent. The partially deployed stent and its delivery system were then withdrawn into the 9-F sheath, and the entire unit was advanced into the shunt and deployed so that the waist of the stent was positioned within the parenchymal tract. Other modifications of this technique, in which sutures are used to constrain the stent, have been reported (37).
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5. Shunt reduction with a constrained stent. Drawing shows a constrained Wallstent placed within a preexisting TIPS. A suture was threaded through the stent mesh and tied with multiple knots to reduce the stent lumen at its midportion. Small curved arrows (here and in the remaining diagrams) indicate turbulent blood flow within the shunt; long arrow shows that overall blood flow through the shunt remains in the antegrade direction.
|
|
A novel alternative approach to constraining a stent for hepatic encephalopathy management was reported by Forauer and McLean (3) (Fig 6). These authors used a balloon-assisted stent to constrain a Wallstent. A 10-mm diameter, 42-mm-long Wallstent was deployed within a 15-mm-long Palmaz stent (P154 M; Johnson & Johnson, Warren, NJ). A 5–0 braided polyglacton suture (Vicryl; Ethicon, Somerville, NJ) was used to tie and anchor the two stents together. The stents were then loaded into a 10-F sheath, deployed into the shunt, and dilated with a 4-mm-diameter, 4-cm-long balloon catheter. The advantage of using a balloon-assisted stent to constrain the Wallstent is the ability to dilate the stent further should complications of portal hypertension recur.
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Shunt reduction with the Forauer-McLean method. (a) Drawing shows a Wallstent placed within a 15-mm Palmaz stent. (b) If necessary, the constrained stent can be expanded in stepwise fashion. Immediately after balloon dilation, the constrained shunt diameter is wider.
|
|
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Shunt reduction with the Forauer-McLean method. (a) Drawing shows a Wallstent placed within a 15-mm Palmaz stent. (b) If necessary, the constrained stent can be expanded in stepwise fashion. Immediately after balloon dilation, the constrained shunt diameter is wider.
|
|
A major limitation of these uncovered constrained stent techniques is the difficulty of accurately regulating blood flow across the shunt and the ability to control and immediately measure the PSG elevation. It is uncertain whether thrombus will form within the dead space surrounding the constrained portion of the stent and alter the flow dynamics sufficiently to reverse hepatic encephalopathy. Patients are usually reexamined several days after the procedure to confirm hemodynamic success.
Adjunct embolization (ie, embolization of dead space surrounding the narrowed portion of the stent) is another development in the evolution of the constrained stent technique. In 1998, Gerbes et al (9) described their experiences with reducing blood flow through a TIPS in three patients. In the first patient, use of an uncovered constrained stent failed to produce changes in flow velocity through the shunt, and hepatic encephalopathy did not resolve. In an attempt to solve this problem, the space between the two stents (the outer TIPS stent and the reducing stent) was filled with an embolic emulsion (Ethibloc; Ethicon, Norderstedt, Germany) (Fig 7). Thereafter, the PSG increased immediately (from 10 to 23 cm H2O), shunt flow velocities decreased to 25 cm/sec (from 85 cm/sec), and hepatic encephalopathy resolved. In the two other cases, the authors placed Palmaz stents into the curved TIPS. To avoid stent displacement, they dilated the distal end of the stent to 6 mm and the proximal end to 10 mm. Flow velocity decreased from 90 to 70 cm/sec as hepatic encephalopathy resolved in one patient. In the other patient, Ethibloc embolization around the Palmaz stent was used and resulted in an immediate increase in PSG. Alternative embolic agents such as Gianturco-Anderson-Wallace coils have been used in a similar fashion (Fig 8).
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Coil embolization of the dead space between inner and outer stents. (a) Drawing illustrates the location of coils used for dead space embolization. (b) Anteroposterior image demonstrates the constrained stent (arrow) within the preexisting TIPS stent (arrowhead). (c) Anteroposterior image shows large coils (arrow) placed between the inner and outer stents. (d) TIPS venogram obtained immediately after coil deployment demonstrates flow throughout the original TIPS lumen. Thrombus has not yet formed within the dead space.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Coil embolization of the dead space between inner and outer stents. (a) Drawing illustrates the location of coils used for dead space embolization. (b) Anteroposterior image demonstrates the constrained stent (arrow) within the preexisting TIPS stent (arrowhead). (c) Anteroposterior image shows large coils (arrow) placed between the inner and outer stents. (d) TIPS venogram obtained immediately after coil deployment demonstrates flow throughout the original TIPS lumen. Thrombus has not yet formed within the dead space.
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. Coil embolization of the dead space between inner and outer stents. (a) Drawing illustrates the location of coils used for dead space embolization. (b) Anteroposterior image demonstrates the constrained stent (arrow) within the preexisting TIPS stent (arrowhead). (c) Anteroposterior image shows large coils (arrow) placed between the inner and outer stents. (d) TIPS venogram obtained immediately after coil deployment demonstrates flow throughout the original TIPS lumen. Thrombus has not yet formed within the dead space.
|
|
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8d. Coil embolization of the dead space between inner and outer stents. (a) Drawing illustrates the location of coils used for dead space embolization. (b) Anteroposterior image demonstrates the constrained stent (arrow) within the preexisting TIPS stent (arrowhead). (c) Anteroposterior image shows large coils (arrow) placed between the inner and outer stents. (d) TIPS venogram obtained immediately after coil deployment demonstrates flow throughout the original TIPS lumen. Thrombus has not yet formed within the dead space.
|
|
Technologic advances have brought constrained stent-grafts to the forefront as an attractive alternative to existing techniques of shunt reduction. Use of a constrained covered stent offers a controlled means of reducing the shunt lumen, with an instantly measurable increase in PSG (Fig 9). Initially, "homemade" endografts were used (Fig 10), but these may be replaced by commercially available stent-grafts.
View larger version (64K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (74K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10c. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10d. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (92K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10e. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10f. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (106K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10g. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
View larger version (120K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10h. Construction of a homemade reduced stent-graft for placement within a TIPS stent. (a) A 4-mm thin-walled e-polytetrafluoroethylene (PTFE) covering was attached to a Palmaz 394 stent. (b) Graft material was sewn to the cross-struts at both ends with a 5-0 prolene suture after the graft was pre-dilated to 6 mm. (c) Stent was crimped on a 6-mm x 8-cm opta-LP (Cordis, Miami Lakes, Fla) balloon and placed through an 11-F peel-away sheath into a 35-cm-long 10-F sheath. (d) A 30-cm-long 12-F sheath (arrows) was then manipulated through the TIPS. (e) Preloaded stent (advanced to the front of the 10-F sheath) (arrows) was then placed so that the leading end of the graft would be near the portal vein entry site. (f, g) Entire device was dilated to 6 mm in this location, and then a 10-mm x 2-cm balloon was used to flair the leading and trailing ends until a seal was obtained. (h) TIPS venogram obtained through the sheath demonstrates the hourglass configuration of the stent-graft, with the dead space (arrows) becoming entirely excluded. Accurate PSG measurements were subsequently made.
|
|
A recent report by Madoff et al (10) described the effective use of constrained stent-grafts (Wallgrafts; Boston Scientific) to treat six patients with refractory hepatic encephalopathy. The placement approach was similar to that described in the initial reports of constrained Wallstents (Fig 11). In their series, either 10- or 12-mm Wallgrafts were fully deployed on the back table. An angioplasty balloon (6- to 8-mm diameter) or dilator (18–24 F) was used as a template to form the endograft waist, with a purse-string 3–0 silk suture (Ethicon) being woven in the stent-graft at approximately one-third the distance from its leading end. The constrained diameter was chosen on the basis of the initial TIPS diameter and the severity of hepatic encephalopathy. A scalpel was used to cut the polyethylene terephthalate (PET) Wallgraft covering from the trailing end of the stent-graft to prevent occlusion of the hepatic vein. The modified Wallgraft was loaded into the tip of a 35-cm-long, 9–10-F sheath and deployed within the existing shunt.
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (178K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (206K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11c. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (222K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11d. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (207K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11e. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11f. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11g. Placement of a constrained Wallgraft endoprosthesis within a preexisting TIPS stent. (a) Anteroposterior image from normal shunt venography through a 9-F vascular sheath shows a TIPS stent. (b) Wallgraft endoprosthesis was deployed on the back table. A dilator was used as a template to determine the reduced endograft diameter. A purse-string suture was woven through the stent mesh and graft material, approximately one-third the distance from the leading end, to create a constrained diameter. (c, d) The trailing end covering of the endograft was removed to prevent occlusion of the hepatic vein following deployment. (e) The stent-graft was then loaded into a new 9-F curved sheath, with the trailing end resheathed first. (f) Anteroposterior image from shunt venography after deployment of the constrained Wallgraft endograft shows an hourglass waist and contrast material within the endograft (white arrow), indicating instantaneous reduction of the shunt diameter. Black arrow indicates the TIPS stent. PSG measurements were obtained. (g) If necessary, an additional constrained endograft (white arrow) may be deployed within the preexisting constrained endograft (black arrow) to further reduce the shunt lumen.
|
|
In the series described by Madoff et al (10), shunt reductions were technically successful in all six patients, with an immediate mean PSG increase of 9.3 mm Hg (8.3–17.6 mm Hg). Clinical improvement (with complete resolution of hepatic encephalopathy in four patients and partial resolution in one patient) was achieved within 72 hours of shunt reduction. The remaining patient continued to decline and died 3 weeks later, at least partially because of a comorbid process, thrombotic thrombocytopenic purpura, which is also known to cause symptoms that mimic hepatic encephalopathy. Shunt occlusion occurred in two patients within 8 months. It is unclear whether the PET graft material played a role in shunt occlusion or whether the final shunt diameters (6 mm) and intimal hyperplasia played a dominant role. In one of these two patients, shunt occlusion occurred after 6 months, leading the authors to favor the latter hypothesis.
Patency rates of various stent-graft materials have been reported in settings of de novo TIPS and in stenotic or previously thrombosed TIPS (38–48). In studies of a porcine model, stents covered with PTFE demonstrated superior patency rates compared with those for the PET stents (47,48). PET stents produced thrombogenic and inflammatory responses that led to earlier shunt occlusion (47,48). In addition, a recent report (49) described three cases in which the PET stents used for TIPS occluded and which were successfully treated with PTFE stent-grafts. From this report, we may infer that PTFE may be superior in situations in which constrained stent-grafts may be needed. However, in the study by Madoff et al (10), the decision to use Wallgrafts was made because of their availability and the ease with which they could be modified, reconstrained, and deployed.
Stent-graft modification by a similar technique was employed successfully to reduce shunt flow, immediately increase PSG, and improve the clinical status of a patient with TIPS-induced hepatic failure (50). In this patient, a PTFE-covered balloon expandable endograft, available currently only outside the United States, was altered by targeted balloon expansion at each end of the device. At the 7-month follow-up, the shunt was patent and liver failure had completely resolved. Although only this single case has been described, modified stent-grafts with PTFE coverings or other minimally porous coverings may be superior, but further study is necessary.
Additional Endovascular Techniques.—
Retrograde embolization of a spontaneous splenorenal shunt is another method of managing hepatic encephalopathy. This technique, which was described in a single case report (51), can be used only when spontaneous splenorenal shunts are present. Such shunts occur in approximately 24% of patients with cirrhosis in whom hepatic encephalopathy presumably is caused by excessive portosystemic shunting.
In the case described by Shioyama et al (51), the patient developed refractory hepatic encephalopathy after TIPS was performed to treat bleeding esophageal varices. The splenorenal shunt was embolized by introducing a 6-F balloon-tipped catheter (20-mm-diameter balloon) from the right internal jugular vein and directing it in retrograde fashion through the left renal vein to the junction of the left renal vein and left suprarenal vein (Fig 12). A coaxially placed microcatheter was used to inject 40 mL of 5% ethanolamine oleate with iopamidol through the balloon catheter after the balloon was inflated. The balloon was held in the splenorenal shunt for approximately 30 minutes. The authors reported that symptoms of hepatic encephalopathy resolved a few days after embolization. Splenic venography performed 2 weeks later showed complete occlusion of the splenorenal shunt. No hepatic encephalopathy recurrence was noted during the next 12 months, and the TIPS was found to be patent at the end of that time. The benefits of this procedure are that the TIPS can remain patent and that further intervention of the TIPS can be performed later, if necessary.
View larger version (67K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12a. Occlusion of a spontaneous splenorenal shunt with the transjugular approach. (a) Drawing illustrates development of a splenorenal shunt with excess portosystemic shunting. Blood flow is diverted toward the left renal vein and into the systemic circulation. (b) Balloon catheter is shown occluding the splenorenal shunt. A microcatheter has been placed for infusion of embolic agent. (c) Following embolization, the splenorenal shunt is completely occluded so that blood now flows through the TIPS and hepatic parenchyma.
|
|
View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12b. Occlusion of a spontaneous splenorenal shunt with the transjugular approach. (a) Drawing illustrates development of a splenorenal shunt with excess portosystemic shunting. Blood flow is diverted toward the left renal vein and into the systemic circulation. (b) Balloon catheter is shown occluding the splenorenal shunt. A microcatheter has been placed for infusion of embolic agent. (c) Following embolization, the splenorenal shunt is completely occluded so that blood now flows through the TIPS and hepatic parenchyma.
|
|
View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12c. Occlusion of a spontaneous splenorenal shunt with the transjugular approach. (a) Drawing illustrates development of a splenorenal shunt with excess portosystemic shunting. Blood flow is diverted toward the left renal vein and into the systemic circulation. (b) Balloon catheter is shown occluding the splenorenal shunt. A microcatheter has been placed for infusion of embolic agent. (c) Following embolization, the splenorenal shunt is completely occluded so that blood now flows through the TIPS and hepatic parenchyma.
|
|
Potential drawbacks of intravascular administration of ethanolamine oleate include severe renal dysfunction (52,53), pulmonary edema (54), disseminated intravascular coagulation (55), and anaphylactic reaction (56,57). Of these complications, renal dysfunction is the most often reported. Theoretically, intravascular hemolysis results in free circulating hemoglobin that is excreted by the glomeruli. The reabsorption of excess amounts of free hemoglobin may result in renal tubular dysfunction. This mechanism may explain why very few reports have described this approach. To reduce the risk of renal failure, haptoglobin, an agent that conjugates hemoglobin, should be administered before the procedure (53).
|
Conclusions
|
|---|
TIPS-related hepatic encephalopathy is a difficult clinical problem that is usually managed conservatively. When hepatic encephalopathy is refractory to these conservative treatments, more invasive techniques are often required. Herein, we described innovative percutaneous interventions that offer options that only recently became available. Of course, these techniques can be improved. Additional study of this complex topic is mandatory so that patients with TIPS-related hepatic encephalopathy can be offered even more ideal treatments.
|
Footnotes
|
|---|
Abbreviations: PET = polyethylene terephthalate,
PSG = portosystemic gradient,
PTFE = polytetrafluoroethylene,
TIPS = transjugular intrahepatic portosystemic shunt
Editor's Note. —In keeping with the highest standards of professional integrity and ethics, RSNA requires that authors of continuing medical educati
Top
Abstract
LEARNING OBJECTIVES FOR TEST...
