Most patients with malignant hilar stenoses are candidates for palliation. For this purpose, biliary drainage plays a major role in improving liver function and managing or avoiding cholangitis. Endoscopic interventions are less invasive than the percutaneous approach and should be considered as the first-line drainage procedures in most cases. Transhepatic interventions should be reserved for endoscopic failures or performed as a complementary approach in a combined procedure. After successful endoscopic access to biliary obstruction, implantation of self-expandable metal stents offers advantages over plastic endoprostheses in terms of stent patency and number of reinterventions.
Causes of malignant biliary obstruction at the level of the hilum can be classified into one of 3 categories ( Box 1 ). The most frequent cause of malignant hilar obstruction is cholangiocarcinoma (CCA), which is the second most common primary hepatobiliary cancer after hepatocellular carcinoma and accounts for 3% of all gastrointestinal cancers worldwide. Approximately 95% of CCAs are adenocarcinomas. CCA is mainly a tumor of the elderly with peak prevalence in the seventh decade of life and affects more men than women. More than 90% of patients present with painless jaundice. CCA-related risk factors are primary sclerosing cholangitis (PSC), choledochal cyst, familial polyposis, hepatolithiasis, congenital hepatic fibrosis, biliary flukes (clonorchiasis, opisthorchiasis), and a history of exposure to thorotrast. A higher prevalence of positive anti–hepatitis C virus antibody test result and hepatolithiasis has been reported to be associated with CCA. There is evidence of a genetic link in the development of CCA. The most common cause of CCA in younger age (third to fifth decade) is PSC, which is located at the hilum in 8% to 40% of the cases. PSC is the most common risk factor in the Western countries. Differentiating between hilar CCA (HCCA) and chronic inflammatory benign hilar strictures in patients with PSC can be extremely difficult.
- 1.
Primary Tumors (cholangiocarcinomas)
- 2.
Local extension (gallbladder cancer, hepatocarcinoma, pancreatic cancer)
- 3.
Lymph node metastases (breast, colon, stomach, ovaries, lymphoma, melanoma)
The most common location of CCA is the main confluence of the hepatic ducts in 60% to 70% of the cases, the distal common bile duct in 20% to 30% cases, and the intrahepatic ducts in 5% to 10% cases. CCA can be anatomically classified as intrahepatic (peripheral), perihilar, or extrahepatic. This classification is important because some of the risk factors are associated with certain locations, for example, hepatolithiasis is more commonly associated with peripheral CCA than with HCCA. The Japanese Liver Cancer Group described different types of tumor growth, with the infiltrating type being the most common at the perihilar region. Infiltrating perihilar lesions are described as Klatskin tumors with an incidence of 1.2 per 100,000 individuals in the United States. Bismuth and Corlette classified malignant hilar stenoses into 4 categories depending on the involvement of the main hepatic ducts and central segmental ducts ( Box 2 ; Fig. 1 ).
- •
Type I: strictures involve the proximal common hepatic duct and spare the confluence of the left and right ductal systems
- •
Type II: strictures involve the confluence and spare the segmental hepatic ducts
- •
Types IIIa and IIIb: strictures involve the right and left segmental hepatic ducts, respectively
- •
Type IV: strictures involve the confluence and both the right and left segmental hepatic ducts
Differentiation between malignant and benign hilar strictures can be difficult but is important for interdisciplinary planning of treatment. The best endoscopic retrograde cholangiography (ERC)-related results can be achieved through the combination of cytologic sampling (eg, biliary brushing) and endoluminal biopsies with a sensitivity of 40% to 60%. Newer cytologic techniques such as digital image analysis and fluorescence in situ hybridization may improve the cytologic accuracy for diagnosing CCA. Endoscopic ultrasound–guided fine-needle aspiration or cholangioscopic targeted biopsies can be useful in selected cases but have not gained wide acceptance.
Because of late clinical symptoms, patients with CCA usually present at an advanced stage of disease and/or have significant comorbidities so that not more than 10% to 20% are candidates for a potential curative surgery. For this reason, palliative treatment plays a major role in the therapy for HCCA. The main aim is to achieve effective biliary decompression for improvement of the liver function and reduction of the risk of cholangitis. For palliation, endoscopic or percutaneous drainage are usually preferred over surgical procedures because they are less invasive. Endoscopic biliary drainage of malignant hilar strictures is often more challenging and complex than management of distal malignant biliary obstruction. In particular, complete drainage of advanced hilar stenoses can be technically very demanding. Nevertheless, endoscopic and percutaneous transhepatic stent implantations have become the standard procedures for palliation of malignant hilar obstruction. In this context, self-expandable metal stents (SEMS) offer several advantages over plastic prostheses. However, the safety and efficacy of their implantation depend on several factors, such as the anatomy of hilar stenoses, the type of stents, the technique of insertion, and the management of technical problems or complications.
Endoscopic versus percutaneous transhepatic approach
Endoscopic and percutaneous techniques allow access to hilar stenoses for subsequent diagnostic and therapeutic procedures. Both methods have advantages and disadvantages. Prophylactic administration of antibiotics is mandatory before any intervention for hilar decompression because of an increased risk of cholangitis, particularly in case of incomplete drainage.
Endoscopic retrograde cholangiopancreatography (ERCP) is recommended as the first-line drainage procedure for the palliation of jaundice in patients with inoperable tumors of Bismuth types I to III. It is less invasive than the percutaneous approach but can be technically more difficult depending on local expertise with both methods. ERCP is associated with a significant risk of cholangitis after contrast injection if there is inadequate drainage of opacified ducts. Therefore, preoperative magnetic resonance cholangiopancreaticography (MRCP) is strongly recommended to plan the strategy for drainage to avoid ducts that cannot be drained. Decompression of the dominant obstructed liver lobe usually provides adequate palliation even in patients with Bismuth type II to III stenoses. Bilateral decompression or drainage of several liver segments can become necessary in high-grade strictures when patients develop cholangitis in undrained segments or if the jaundice fails to resolve after incomplete decompression.
The percutaneous approach is more invasive than ERC. Percutaneous transhepatic cholangiography (PTC) is indicated when MRCP reveals that appropriate biliary drainage by means of ERCP can most likely not be obtained or if endoscopic drainage had failed, particularly in patients with Klatskin tumor of Bismuth type III or IV. In patients with biliary obstruction, PTC should be followed by percutaneous transhepatic cholangiographic drainage (PTCD) to achieve decompression of the biliary tree and to reduce the risk of cholangitis. PTCD and can be combined with percutaneous transhepatic cholangioscopy (PTCS), photodynamic therapy (PDT), and biliary stent placement. PTCS can be useful for differentiation of benign and malignant hilar strictures and insertion of guidewires under direct visual control.
Another possibility for biliary access in a difficult anatomy is to combine both methods in a rendezvous maneuver. For this approach, a guidewire is inserted transhepatically and grasped endoscopically with a basket catheter or a snare. This method achieves an endoscopic approach to the biliary tract without the need for further percutaneous interventions and can be particularly helpful in advanced stages of hilar stenoses when there is a need for placement of several stents.
Successful initial drainage, regardless of the procedure, is the most important factor in determining the clinical outcome. Retrospective studies show that ERCP-related mortality increases if hepatic segments have been opacified without subsequent drainage. Therefore, rescue percutaneous drainage should be considered in case of failed endoscopic decompression of segments that have been opacified.
The selection criteria for endoscopic or percutaneous drainage depend on anatomic factors determined by MRCP, the estimated number of stents required for appropriate biliary decompression, and local experience. A recent retrospective comparison between endoscopic and percutaneous implantation of SEMS for Bismuth type III and IV stenoses demonstrated a significantly higher success rate for PTCD. There was no significant difference in terms of the incidence of cholangitis, overall complications, procedure-related mortality, and stent patency. Survival was significantly longer in those with successful drainage independent of the type of intervention.
Types of stent
Biliary decompression of hilar obstruction can be obtained with plastic stents or SEMS. Straight, slightly curved, or pigtail stents are the most commonly used types of plastic endoprostheses. They are less expensive than metal stents but have a higher risk of an early occlusion with the consequence of recurrent jaundice and cholangitis. If signs of cholangitis develop, stent replacement is necessary to avoid the development of life-threatening sepsis. Stent occlusion is associated with additional days of hospitalization and antibiotic therapy, and a higher number of reinterventions. Attempts to increase the patency of plastic stents, such as by administration of antibiotics and/or ursodeoxycholic acid and by the use of different materials and special coatings, do not lead to significant improvement in patency. Plastic stent patency depends not only on the material but also on the diameter, which is limited to 12F because larger devices cannot be inserted through the instrumentation channel of a standard therapeutic duodenoscope. Several uncontrolled studies have shown that SEMS have a lower obstruction rate than plastic stents in patients with proximal common bile duct stenosis. The median patency of metal stents was 8.9 months for patients with primary bile duct tumors and 5.4 months for all patients, and this advantage was not related to the Bismuth classification. SEMS are cost-effective for patients who survive for at least 3 to 6 months. Therefore metal stents should be considered for patients with a longer survival expectancy. The most common causes of metal stent occlusion are growth of hyperplastic and/or tumor tissue through the lattice of the metal mesh and tumor overgrowth.
A variety of uncovered SEMS are commercially available. They mainly differ in size and material that the wire mesh is composed of. The effect of these parameters on clinical outcome is difficult to determine because of a lack of formal comparative trials. Predeployed SEMS are mounted on 6F to 8F catheters. After positioning and release, the stent diameter increases to 8 to 10 mm. Over time, the open mesh is subsequently covered by biliary epithelial cells. In contrast to plastic stents, repositioning or removal of uncovered SEMS is difficult or impossible once they have been deployed, which is why SEMS should only be used in patients with proven malignancy and unresectable tumors or in nonoperative patients. Fully covered SEMS are not indicated for proximal biliary obstruction because they may occlude the contralateral biliary system or biliary side branches. When fully expanded, nearly all commercially available SEMS have diameters of 8 or 10 mm and lengths between 4 and 10 cm. The selection depends on ductal diameter, number of SEMS required, and length of stenoses. The Wallstent (Boston Scientific, Natick, MA, USA) is a metal stent consisting of a stainless steel alloy tubular mesh. The stent length on the delivery catheter decreases by about 30% (foreshortening) after release, which has to be considered for correct positioning. The biliary Wallflex stent (Boston Scientific) is a braided platinum core nitinol stent with less shortening than the Wallstent ( Fig. 2 ). The Zilver stent (Cook Medical Inc, Bloomington, IN, USA) is also made from nitinol with a laser-cut Z-like arrangement of the meshes ( Fig. 3 ). It is nonforeshortening, which facilitates precise positioning. This stent has outcomes comparable with that of Wallstent for palliative drainage of distal malignant biliary stenoses with respect to technical success, occlusion rates, and overall patency. Nitinol stents conform better to the biliary anatomy than stainless steel SEMS, which is particularly important in an angulated ductal system or a tortuous anatomy. In addition to this potential advantage, a subgroup analysis of a retrospective review in 101 patients showed a significantly longer stent patency in malignant hilar stenosis with a Niti-D biliary (Taewoong Medical Co Ltd) uncovered stent as opposed to the uncovered Wallstent. To date, prospective randomized controlled studies of different types of SEMS for malignant hilar obstruction have not been reported. Several strategies for prolongation of metal stent patency are currently under investigation. A pilot study indicated that the endoscopic insertion of a metal stent covered with a paclitaxel-incorporated membrane is technically feasible, safe, and effective in patients with malignant biliary obstruction. Paclitaxel can suppress tissue reactions to metal stents. Further studies are necessary to evaluate the adequate drug dose that exerts an antitumorous effect without damaging the adjacent normal biliary mucosa.
Types of stent
Biliary decompression of hilar obstruction can be obtained with plastic stents or SEMS. Straight, slightly curved, or pigtail stents are the most commonly used types of plastic endoprostheses. They are less expensive than metal stents but have a higher risk of an early occlusion with the consequence of recurrent jaundice and cholangitis. If signs of cholangitis develop, stent replacement is necessary to avoid the development of life-threatening sepsis. Stent occlusion is associated with additional days of hospitalization and antibiotic therapy, and a higher number of reinterventions. Attempts to increase the patency of plastic stents, such as by administration of antibiotics and/or ursodeoxycholic acid and by the use of different materials and special coatings, do not lead to significant improvement in patency. Plastic stent patency depends not only on the material but also on the diameter, which is limited to 12F because larger devices cannot be inserted through the instrumentation channel of a standard therapeutic duodenoscope. Several uncontrolled studies have shown that SEMS have a lower obstruction rate than plastic stents in patients with proximal common bile duct stenosis. The median patency of metal stents was 8.9 months for patients with primary bile duct tumors and 5.4 months for all patients, and this advantage was not related to the Bismuth classification. SEMS are cost-effective for patients who survive for at least 3 to 6 months. Therefore metal stents should be considered for patients with a longer survival expectancy. The most common causes of metal stent occlusion are growth of hyperplastic and/or tumor tissue through the lattice of the metal mesh and tumor overgrowth.
A variety of uncovered SEMS are commercially available. They mainly differ in size and material that the wire mesh is composed of. The effect of these parameters on clinical outcome is difficult to determine because of a lack of formal comparative trials. Predeployed SEMS are mounted on 6F to 8F catheters. After positioning and release, the stent diameter increases to 8 to 10 mm. Over time, the open mesh is subsequently covered by biliary epithelial cells. In contrast to plastic stents, repositioning or removal of uncovered SEMS is difficult or impossible once they have been deployed, which is why SEMS should only be used in patients with proven malignancy and unresectable tumors or in nonoperative patients. Fully covered SEMS are not indicated for proximal biliary obstruction because they may occlude the contralateral biliary system or biliary side branches. When fully expanded, nearly all commercially available SEMS have diameters of 8 or 10 mm and lengths between 4 and 10 cm. The selection depends on ductal diameter, number of SEMS required, and length of stenoses. The Wallstent (Boston Scientific, Natick, MA, USA) is a metal stent consisting of a stainless steel alloy tubular mesh. The stent length on the delivery catheter decreases by about 30% (foreshortening) after release, which has to be considered for correct positioning. The biliary Wallflex stent (Boston Scientific) is a braided platinum core nitinol stent with less shortening than the Wallstent ( Fig. 2 ). The Zilver stent (Cook Medical Inc, Bloomington, IN, USA) is also made from nitinol with a laser-cut Z-like arrangement of the meshes ( Fig. 3 ). It is nonforeshortening, which facilitates precise positioning. This stent has outcomes comparable with that of Wallstent for palliative drainage of distal malignant biliary stenoses with respect to technical success, occlusion rates, and overall patency. Nitinol stents conform better to the biliary anatomy than stainless steel SEMS, which is particularly important in an angulated ductal system or a tortuous anatomy. In addition to this potential advantage, a subgroup analysis of a retrospective review in 101 patients showed a significantly longer stent patency in malignant hilar stenosis with a Niti-D biliary (Taewoong Medical Co Ltd) uncovered stent as opposed to the uncovered Wallstent. To date, prospective randomized controlled studies of different types of SEMS for malignant hilar obstruction have not been reported. Several strategies for prolongation of metal stent patency are currently under investigation. A pilot study indicated that the endoscopic insertion of a metal stent covered with a paclitaxel-incorporated membrane is technically feasible, safe, and effective in patients with malignant biliary obstruction. Paclitaxel can suppress tissue reactions to metal stents. Further studies are necessary to evaluate the adequate drug dose that exerts an antitumorous effect without damaging the adjacent normal biliary mucosa.
Techniques of implantation and clinical results
There is controversy as to whether complete drainage of hilar obstruction is necessary. Preoperative MRCP should be performed in all patients with suspected proximal biliary stenoses. MRCP and magnetic resonance imaging (MRI) of the liver allow classification of hilar stenoses according to the Bismuth categories and determination of the volume of obstructed liver segments. In addition, it provides further information on tumor staging in terms of metastases and tumor involvement of vessels. This information is mandatory for interdisciplinary decision making in view of tumor resectability or palliation, including biliary drainage and adjunctive treatment such as PDT and/or radiochemotherapy. Palliative drainage of a Bismuth type I stricture can be achieved by a single stent because both main hepatic ducts are not involved and communicate with each other. In patients with Bismuth types II to IV, MRCP is helpful to decide which obstructed liver segments should be drained and how many stents may be needed. The endoscopic or percutaneous route should be selected depending on the anatomy and the local experience with these techniques. In most cases of Bismuth type II stenoses, a single stent placement into the right or left hepatic system is sufficient. The selection of the side depends on the volume of the obstructed liver lobe and presence or absence of atrophy. In Bismuth types III to IV obstructions, more than 1 stent can be required to achieve relief from jaundice. A retrospective study in 107 patients showed that drainage of more than 50% of the liver volume seems to be an important predictor of clinical effectiveness especially in Bismuth type III strictures. Recent studies suggest that the higher the degree of stenoses the more is it likely that implantation of a single stent will not achieve effective palliative drainage. Endoscopic placement of 2 or more SEMS can be very difficult and complex. In any case, it is extremely important to avoid opacification of segments with contrast media upstream from the level of stenosis without achieving subsequent drainage because of an increased risk of cholangitis and a higher mortality rate. This risk can be reduced by insertion of a guidewire toward the targeted liver lobe or an obstructed segment under fluoroscopic guidance without contrast injection. A catheter is then advanced over the wire for subsequent injection of contrast medium. After appropriate opacification showing details of the anatomy of the obstructed segment, the catheter is removed, and the guidewire remains in place. Depending on the MRCP findings and need for complete drainage, 2 or more guidewires can be placed simultaneously with the same technique. When more than 1 stent is placed, all strictures should be balloon dilated to 6 mm to facilitate placement of stents and allow a rapid expansion. The technique of implantation of more than 1 stent depends on the type of SEMS. Wallstents or Wallflex stents can be sequentially inserted and released because the smooth outer surface allows insertion of further stents alongside a released stent ( Fig. 4 ). This technique should not be tried with Zilver stents because the laser-cut metallic filaments of a delivered stent may prevent advancement of another stent alongside of it. However the 6F diameter delivery catheters allow the parallel insertion of 2 predeployed stents through the working channel of a therapeutic endoscope. After correct positioning, 2 stents can then be stepwise and simultaneously released ( Fig. 5 ). In patients in whom hilar stenoses do not involve the common bile duct, SEMS do not need to extend through the papilla (see Fig. 4 ). However, reinterventions may become more difficult with complete intraductal positioning. For this purpose, it is important to position the distal ends of stents next to each other so that each lumen can be individually cannulated in case of need of reintervention (see Fig. 5 ). An alternative to side-by-side stent placement is the stent-through-stent technique. It was initially described when using the Wallstent, but it can be very difficult using these stents because of the tight wire mesh. In contrast, the recently developed dedicated nitinol stents have a widened mesh structure in the midsection of the stent, which facilitates insertion of a second guidewire through this section, such as into the contralateral liver lobe or another obstructed segment followed by implantation of a second stent ( Fig. 6 ).
Related posts:
Pancreaticobiliary and Gastrointestinal Stents
Expandable Metal Stents: Principles and Tissue Responses
Expandable Stents for Benign Esophageal Disease
Plastic Biliary Stents for Malignant Biliary Diseases
Expandable Metal Stents for Benign Biliary Disease
Self-Expandable Metal Stents for Malignant Distal Biliary Strictures
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
