U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Kufe DW, Pollock RE, Weichselbaum RR, et al., editors. Holland-Frei Cancer Medicine. 6th edition. Hamilton (ON): BC Decker; 2003.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of Holland-Frei Cancer Medicine

Holland-Frei Cancer Medicine. 6th edition.

Show details

Bile Duct Cancer

, MD, FACS.

Cholangiocarcinomas are malignant tumors that arise from the epithelium of the intrahepatic or extrahepatic bile ducts. Cholangiocarcinomas are rare compared to hepatocellular carcinoma, comprising less than 10% of primary malignancies of the liver.97 In the United States, approximately 3,000 patients are diagnosed with cholangiocarcinoma of both the intra- or extrahepatic biliary system annually.98 The autopsy prevalence of cholangiocarcinoma is low also, being reported in 0.089% to 0.46% of necropsies.97 Cholangiocarcinomas are diagnosed most frequently in the fifth and sixth decades of life.99 There is only a slight male preponderance of cases of cholangiocarcinoma. Cholangiocarcinomas can arise at any site in the intra- or extrahepatic biliary system, but perihilar tumors comprise two-thirds of the cases of cholangiocarcinoma (Figure 102-5).99

Figure 102-5. The distribution of 294 cholangiocarcinomas into intrahepatic, perihilar, and distal subgroups.

Figure 102-5

The distribution of 294 cholangiocarcinomas into intrahepatic, perihilar, and distal subgroups. Reprinted with permission from Nakeeb et al.

Causative Factors

There are distinct differences between the factors associated with cholangiocarcinoma and those associated with hepatocellular carcinoma (HCC) (Table 102-2). Cholangiocarcinoma does not appear to be associated with hepatitis B or C virus infection, or mycotoxin exposure.97 Only 10% to 20% of cholangiocarcinomas occur in cirrhotic patients, compared to the 70 to 90% of hepatocellular carcinomas that arise in cirrhotic livers.98,100,101 A cohort study in Denmark of 11,605 patients with cirrhosis indicated a 60-fold increased risk for developing hepatocellular cancer and a tenfold increased risk for cholangiocarcinoma.101 Frequently, the cirrhosis associated with cholangiocarcinomas is a subacute secondary biliary type that results from the neoplastic obstruction of the bile ducts, indicating that, in some cases, cirrhosis in cholangiocarcinoma patients is the result of the tumor rather than its cause.

Table 102-2. Factors Associated with an Increased Risk to Develop Cholangiocarcinoma Compared with Factors Associated with Hepatocellular Carcinoma.

Table 102-2

Factors Associated with an Increased Risk to Develop Cholangiocarcinoma Compared with Factors Associated with Hepatocellular Carcinoma.

Cholangiocarcinoma is more prevalent in Southeast Asia than in other parts of the world. The higher incidence in this geographic region is related to parasitic infection with the liver flukes Opisthorchis sinensis and O. viverrini.102,103 Infestation by other biliary parasites, such as Fasciola hepatica and Schistosomiasis japonica, does not appear to have similar carcinogenic effects.104 Liver flukes induce hyperplasia, fibrosis, and adenomatous proliferation of human biliary epithelium and are associated with hepatolithiasis.105,106 The fluke infestation suggests a direct etiologic role in the subsequent development of cholangiocarcinoma, but this relationship is not established unequivocally. It is interesting to note that in a Syrian golden hamster model, administration of dimethylnitrosamine alone does not cause neoplasia, but this agent given to animals with liver fluke infection leads to the development of cholangiocarcinomas.107,108

Several disorders that can produce chronic inflammation of the bile ducts have been associated with an increased risk of developing cholangiocarcinoma. These include polycystic liver disease, choledochal cysts, congenital dilation of the intrahepatic bile ducts (Caroli syndrome), sclerosing cholangitis (occasionally in association with inflammatory bowel disease), hepatolithiasis, and cholelithiasis.109–118 Hepatolithiasis is not a common disorder, and only 5% to 7% of patients with documented hepatic stones develop cholangiocarcinoma.117,118 The reported incidence of cholangiocarcinoma developing in areas of congenital cystic dilation of the bile duct, including choledochal cysts and Caroli disease, ranges from 3% to 30%.119,120 Patients with primary sclerosing cholangitis are also at increased risk to develop cholangiocarcinoma, with incidence rates ranging from 9% to 40%.121–123 Patients with ulcerative colitis may also develop sclerosing cholangitis, but cholangiocarcinoma occurs in only 0.4% to 1.4% of individuals with ulcerative colitis.121 In patients with sclerosing cholangitis, whether associated with ulcerative colitis or not, radiologic distinction between sclerosing cholangitis and cholangiocarcinoma is often impossible. A recent study showed that the serum tumor marker CA19-9 had an 89% sensitivity and 86% specificity in diagnosing cholangiocarcinoma in patients with sclerosing cholangitis.124 Combining serum CA19-9 levels with serum CEA levels may further increase the diagnostic accuracy to detect cholangiocarcinoma in patients with sclerosing cholangitis.125

Patients who underwent diagnostic radiography with intravenous injection of Thorotrast (thorium dioxide) are at high risk of developing hepatocellular carcinoma, angiosarcoma, and cholangiocarcinoma.126,127 Cholangiocarcinoma is the most frequent hepatic neoplasm reported in patients who have received Thorotrast. Exposure to several drugs or carcinogens has also been linked to an increased risk to develop cholangiocarcinoma (see Table 102-2). Because cholangiocarcinoma is a relatively rare neoplasm, it has been difficult to prove its pathogenesis related to any of these factors, but it is clear that chronic inflammation of the biliary tree by any cause is associated with an increased risk of developing cholangiocarcinoma.

It has been suggested that HCC and cholangiocarcinoma arise from a common hepatic progenitor, or “stem” cell, in response to chronic injury and subsequent cellular proliferation.128 Evidence supporting this theory occurs in patients who present with combined HCC and cholangiocarcinoma.129,130 Approximately 3% of patients with HCC have histologic evidence within or adjacent to the cancer of coexistent cholangiocarcinoma.130 Evaluation of a variety of cell surface markers in the cholangiocarcinoma and HCC cells of patients with combined tumors indicates a high degree of homology in the expression of cell surface molecules between the two malignant cell types.129,130 Furthermore, the only consistent genetic alterations that have been detected in cholangiocarcinomas are loss of heterozygosity in regions of chromosomes 5 and 17 that are also lost in hepatocellular carcinoma.131 These findings suggest a common origin from a pluripotential cell type in patients with combined HCC and cholangiocarcinoma, but definitive proof has not yet been provided.

Chronic inflammation of the biliary system or exposure to genotoxic agents concentrated in bile may produce damage to the DNA of biliary epithelial cells, leading to the development of cholangiocarcinoma. Mutations in the p53 tumor suppressor gene and in the K-ras protooncogene have been identified in cholangiocarcinoma patients.132–134 There may be geographic and population-based differences in the mutation rates of these two genes in cholangiocarcinoma, but alterations in p53 and K-ras are observed in significant proportions of patients with any of the identified factors (see Table 102-2) that increase risk to develop cholangiocarcinoma. Overexpression of c-erbB-2, a protooncogene that encodes a transmembrane protein which is highly homologous to epidermal growth factor receptor, has been confirmed in human cholangiocarcinoma cells and in benign proliferative biliary epithelium from patients with hepatolithiasis, primary sclerosing cholangitis, and liver fluke infestation.135 Alterations in c-erbB-2 expression may occur early in the chronic inflammation-induced proliferation of biliary epithelium leading to malignant transformation. Chronic inflammation may also produce the overexpression of the Bcl-2 protooncogene observed in cholangiocarcinomas, which may promote tumorigenesis by inhibiting normal apoptotic processes.136

Clinical Presentation

The clinical features of cholangiocarcinoma are nonspecific and depend on the location of the tumor. The usual clinical presentation of patients with hilar cholangiocarcinoma is painless jaundice. Patients may also report concomitant onset of fatigue, pruritus, fever, vague abdominal pain, and anorexia. The serum liver function tests in patients with hilar cholangiocarcinoma commonly demonstrate obstructive jaundice, with alkaline phosphatase and total bilirubin levels elevated in greater than 90% of patients.98 Cholangiocarcinomas that arise in peripheral bile ducts within the hepatic parenchyma usually reach a large size before becoming clinically evident. Patients with these large peripheral hepatic tumors usually present with hepatomegaly and an upper abdominal mass, abdominal and back pain, and weight loss.98 Jaundice and ascites are late and usually preterminal sequelae in patients with large intrahepatic cholangiocarcinomas. Jaundice associated with a large hepatic cholangiocarcinoma is caused by a combination of extension of the tumor to the bifurcation of the left and right hepatic ducts, and by compression of contralateral bile ducts by the expanding tumor.

Serum alkaline phosphatase levels are elevated in more than 90% of patients with cholangiocarcinoma.99 Serum bilirubin also is elevated in the majority of cholangiocarcinoma patients, particularly in those with a tumor arising in the central portion of the liver or the extrahepatic hilar bile ducts.137 In contrast to hepatocellular carcinoma, serum α-fetoprotein levels are abnormal in fewer than 5% of cholangiocarcinoma patients.99 There is an increase in serum CEA levels in 40% to 60% of cholangiocarcinoma patients.99,138 Another tumor marker, CA19-9, is elevated in more than 80% of patients with cholangiocarcinoma.138 Mild anemia occurs occasionally, but other serum laboratory studies are usually normal. However, clinically significant hypercalcemia in the absence of bone metastases has been described in several patients with large hepatic cholangiocarcinomas.139 In one case, the hypercalcemia was shown to be caused by release of parathyroid hormone-related protein by the cholangiocarcinoma.140

Pathology

Cholangiocarcinomas appear grossly as firm, gray-white tumors. Those originating in the periphery of the hepatic parenchyma usually are solitary and large, but satellite nodules occasionally are present.97,141 Gross tumor invasion of the large portal or hepatic veins occurs much less frequently than in hepatocellular carcinoma. The gross and microscopic appearance of intrahepatic cholangiocarcinomas may have prognostic significance because tumors with periductal infiltration have a higher incidence of lymph node and intrahepatic metastasis.142 Metastases to the regional lymph nodes, lungs, and peritoneal cavity are more common in cholangiocarcinoma than in hepatocellular carcinoma. When the tumor causes long-standing biliary obstruction, the liver may show secondary biliary cirrhosis.

Microscopically, cholangiocarcinoma is characterized by low cuboidal cells that resemble the normal biliary epithelium. Varying degrees of pleomorphism, atypia, mitotic activity, hyperchromatic nuclei, and prominent nucleoli are noted from area to area in the same tumor. In more poorly differentiated tumors, solid cords of cells without lumens may be present. Rarely, a clear cell variant of cholangiocarcinoma occurs, which must be distinguished from clear cell renal carcinoma with liver metastasis.143 Cholangiocarcinomas are mucin-secreting adenocarcinomas, and intracellular and intraluminal mucin often can be demonstrated. The presence of mucin is useful in differentiating cholangiocarcinoma from hepatocellular carcinoma. The absence of bile production by cholangiocarcinoma can also be useful in distinguishing this tumor from a hepatocellular carcinoma. Immunohistochemical staining that is positive for epithelial membrane antigen and tissue polypeptide antigen may be useful in confirming a diagnosis of cholangiocarcinoma.144,145 Immunohistochemical staining for cytokeratin subtypes can be helpful in differentiating cholangiocarcinoma from metastatic colorectal carcinoma.146 Cholangiocarcinomas are usually locally invasive, with spread along nerves or in subepithelial layers of the bile ducts.

Diagnostic Studies

Peripheral intrahepatic cholangiocarcinoma is often difficult to distinguish pathologically and radiographically from a deposit of metastatic adenocarcinoma within the liver. While transabdominal ultrasonography can detect an intrahepatic malignant tumor greater than 2 cm in diameter, ultrasound findings do not differ between cholangiocarcinomas, liver metastases from extrahepatic adenocarcinomas, and multinodular hepatocellular carcinoma.147 Computed tomography (CT) demonstrates a rounded, low attenuation mass with irregular or lobulated margins (Figure 102-6). Satellite lesions may be evident, particularly when using helical CT during the optimal period of hepatic contrast enhancement. Calcification within the tumor is present in 25% of cases, and a central scar is observed in 30%.148 Magnetic resonance imaging (MRI) shows a nonencapsulated mass with irregular margins that is hypointense compared to the normal liver on T1-weighted and hyperintense on T2-weighted images. The peripheral rim of the tumor usually enhances following MRI contrast administration. A hyperintense central scar is best seen on T2-weighted images, but the CT and MRI characteristics of intrahepatic cholangiocarcinomas may be present in other types of hepatic tumors.148

Figure 102-6. High-resolution, helical CT scan during the arterial contrast phase in a patient with an intrahepatic cholangiocarcinoma.

Figure 102-6

High-resolution, helical CT scan during the arterial contrast phase in a patient with an intrahepatic cholangiocarcinoma. The periphery of the tumor (arrow) has irregular margins and enhances with contrast. A relatively hypovascular area of scar and tumor (more...)

The diagnosis of pancreatic cancer is considered frequently in patients presenting with painless jaundice. For this reason, a CT scan of the abdomen may be the first radiologic study obtained. Patients with painless jaundice due to pancreatic cancer may or may not have a mass in the head of the pancreas on a CT scan. They will, however, have dilation of the extrahepatic biliary tree and gallbladder (if the latter has not been removed previously). In contrast, a diagnosis of hilar cholangiocarcinoma should be suspected in the patient with painless jaundice whose CT scan demonstrates dilated intrahepatic bile ducts with a normal gallbladder and extrahepatic biliary tree. High-resolution, helical CT scans can provide information on the location of an obstructing biliary tumor and may suggest the extent of involvement of the liver and porta hepatis structures by the tumor (Figures 102-7, 102-8). Multiphasic helical CT can correctly identify the level of biliary obstruction by a hilar cholangiocarcinoma in 63% to 90% of patients.149–151 Preoperative helical CT is also useful in demonstrating lobar or segmental liver atrophy caused by bile duct obstruction or portal vein occlusion (Figure 102-9).151 However, helical CT is not accurate in assessing the resectability of hilar cholangiocarcinomas because of limited resolution in evaluating intraductal tumor spread and significant false-positive and false-negative rates in demonstrating portal vein or hepatic artery involvement by tumor.149–151

Figure 102-7. High-resolution, helical CT scan in a patient with obstructive jaundice.

Figure 102-7

High-resolution, helical CT scan in a patient with obstructive jaundice. The scan demonstrates a tumor (large arrow) anterior to the portal vein with a stent (white area adjacent to large arrow) in place. Dilated intrahepatic bile ducts (small arrows (more...)

Figure 102-8. High-resolution, helical CT scan in another patient presenting with obstructive jaundice.

Figure 102-8

High-resolution, helical CT scan in another patient presenting with obstructive jaundice. The tumor mass (large arrow, left panel), producing marked intrahepatic biliary duct dilatation is evident. Areas of tumor invasion of the portal vein (small arrows, (more...)

Figure 102-9. High-resolution, helical CT scan in a patient presenting with several months of increasing pruritus followed by the development of clinically evident jaundice.

Figure 102-9

High-resolution, helical CT scan in a patient presenting with several months of increasing pruritus followed by the development of clinically evident jaundice. The relatively hypodense hilar cholangiocarcinoma (large arrow) is evident. Marked atrophy (more...)

Ultrasonography is the simplest noninvasive study for the jaundiced patient. Like the CT scan, ultrasonography can demonstrate a nondilated gallbladder and common bile duct associated with dilated intrahepatic ducts. In addition, as gray-scale ultrasonography has improved, the diagnosis of cholangiocarcinoma is supported by finding a hilar bile duct mass in 65% to 90% of patients.152,153 Ultrasonography and CT scan may be used to demonstrate the presence of intrahepatic tumor due to direct extension or noncontiguous metastases and enlarged periportal lymph nodes, suggesting nodal metastases.153,154 Even intraoperative ultrasonography is suboptimal for detecting intraductal spread by hilar cholangiocarcinoma, correctly demonstrating the extent of tumor spread away from the primary biliary tumor in only 18% of cases.155 Intraoperative ultrasonography can be used to screen for noncontiguous liver metastases from the primary biliary cancer and can accurately detect direct tumor invasion of the portal vein or hepatic artery in 83.3% and 60% of cases, respectively.155 Recently, endoscopic ultrasonography and intraductal sonography findings have been described in patients with bile duct cancer, but the small number of patients studied precludes determination of the staging accuracy of these techniques.156

Similar to the intrahepatic variety, hilar cholangiocarcinoma usually shows hypointensity on T1- and hyperintensity on T2-weighted MRI. Dilated intrahepatic bile ducts are evident in patients with obstructing tumors, and lobar atrophy is seen in cases of portal venous occlusion. Fast low-angle shot (FLASH) MR with contrast-enhanced coronal imaging has been used to demonstrate intraluminal extension of tumor and to distinguish between blood vessels and bile ducts.157,158 Magnetic resonance cholangiopancreatography (MRCP) and MR virtual endoscopy can demonstrate hilar bile duct obstruction by tumor with dilated intrahepatic ducts.157,159 The advantages of MRCP over direct cholangiography include noninvasiveness and possible visualization of isolated bile ducts. However, MRCP may have limitations relative to direct cholangiography because evaluation of tumor extent is limited by spatial resolution.159

Cholangiography definitively demonstrates a lesion obstructing the left and right hepatic duct at the hilar confluence (Figure 102-10), and percutaneous transhepatic cholangiography (PTC) and endoscopic retrograde cholangiopancreatography (ERCP) are both useful in assessing patients with extrahepatic biliary obstruction. A prospective, randomized comparison of PTC and ERCP in patients with jaundice concluded that the two techniques had similar diagnostic accuracy.160 PTC was 100% accurate at demonstrating obstruction at the confluence of the left and right hepatic ducts, while ERCP had an accuracy of 92% in demonstrating these lesions. ERCP has the additional benefit of providing a pancreatogram. A normal pancreatogram helps to exclude a small carcinoma of the head of the pancreas as a cause of biliary obstruction. Some investigators have recommended combined PTC and ERCP to establish the extent of the lesion in the bile ducts; however, such concomitant studies are helpful only in selected patients with complete obstruction of the biliary tree.156,161 Cytologic specimens can be obtained at the time of PTC and ERCP. The presence of malignant cells in bile or bile duct brushings is confirmed in approximately 50% of patients undergoing PTC or ERCP.152,156,160,161

Figure 102-10. Endoscopic retrograde cholangiopancreatography (ERCP) showing a focal stricture of the proper hepatic bile duct (arrow) with marked dilatation of the intrahepatic bile ducts.

Figure 102-10

Endoscopic retrograde cholangiopancreatography (ERCP) showing a focal stricture of the proper hepatic bile duct (arrow) with marked dilatation of the intrahepatic bile ducts. This hilar cholangiocarcinoma was completely resected with Roux-Y hepaticojejunostomy (more...)

Drainage of the obstructed biliary tree with partial or complete relief of jaundice and associated symptoms can be achieved with PTC. Improvements in catheter technology led to the development of endoprostheses that can be placed across the malignant obstruction into the duodenum to allow internal drainage.162 It must be emphasized that providing symptomatic relief for patients by decompressing the biliary tract should not be the primary reason to place these catheters. Prospective, randomized studies have failed to demonstrate a benefit in terms of a decrease in hospital morbidity or mortality by preoperative decompression of biliary obstruction.163,164 However, the catheters are useful in identifying and dissecting the hepatic duct bifurcation at the time of operation and aid in the reconstruction of the biliary tract following extirpation of the tumor.165,166 Although ERCP can be employed to place an internal stent across a malignant hilar obstruction, the success rate with this procedure is much lower than that with PTC.167

Positron emission tomography (PET) is being evaluated as a diagnostic tool in patients with all types of malignant tumors. PET assesses in vivo metabolism of positron-emitting radiolabeled tracers like [18F] fluoro-2-deoxy-D-glucose (18F-FDG), a glucose analog that accumulates in various malignant tumors because of their high glucose metabolic rates. FDG-PET does not provide anatomic detail to assess resectability of hilar cholangiocarcinomas or intrahepatic malignancies, but it may prove useful in detecting distant metastatic disease that would preclude a curative resection. In patients with primary sclerosing cholangitis, FDG-PET studies may be able to detect small hilar and intrahepatic cholangiocarcinomas and thus may be useful in therapeutic and transplant decision making in these patients.168 PET has recently been used to aid in the diagnosis and staging of patients with bile duct cancer.169 PET scan images correctly detected the primary cholangiocarcinoma in 24 of 26 patients (sensitivity 92.3%) and was true negative in 8 patients with benign bile duct disease (adenoma, sclerosing cholangitis, Caroli disease). Distant metastatic disease was diagnosed correctly in 7 of 10 patients with histologically proven metastases, but regional lymph node metastases were identified in only 2 of 15 patients (13.3%).169

The role of laparoscopy as part of the diagnostic and staging evaluation of patients with hilar cholangiocarcinoma is being evaluated at our institution. Several patients with seemingly resectable tumors have avoided an exploratory laparotomy when peritoneal tumor implants were found with laparoscopy. In addition, patients at high risk of developing peritoneal carcinomatosis may be identified by positive cytologic specimens obtained from laparoscopic washings. Finally, laparoscopic ultrasonography can be used to exclude the presence of noncontiguous liver metastases or extensive hilar tumor infiltration in patients with extrahepatic bile duct cancers.170

Treatment of Intrahepatic Cholangiocarcinoma

The majority of patients with intrahepatic cholangiocarcinoma present with large tumors and usually have evidence of regional lymph node, pulmonary, and/or bone metastases at the time of diagnosis. In patients who present with jaundice from large intrahepatic cholangiocarcinomas, death usually ensues within a year of diagnosis. Patients with an elevated serum bilirubin level associated with an intrahepatic cholangiocarcinoma will rarely be candidates for an attempt at curative resection because of coexistent hepatic artery and portal vein invasion, extensive lymph node metastases, bilobar liver involvement by tumor, and/or distant metastases.171 Intrahepatic cholangiocarcinomas may be detected in 30% to 45% of patients before they metastasize or cause jaundice.171,172 These patients should be considered for operation because long-term survival has been reported in a proportion of the patients undergoing curative liver resection for intrahepatic cholangiocarcinoma.100,171–177 A study of 19 patients who underwent resection of intrahepatic cholangiocarcinoma demonstrated that patients with no porta hepatis lymph node metastases had a 3-year survival rate of 64% compared to 0% for patients with nodal metastases.176 A larger cohort of 32 patients who underwent resection of intrahepatic cholangiocarcinomas confirmed the negative prognostic impact of regional lymph node metastases and large size (> 5 cm diameter) of the primary tumor.177 The 5-year overall survival rates reported for patients who underwent a margin-negative liver resection for intrahepatic cholangiocarcinoma range from 20% to 48%, with regional lymph node metastases, presence of satellite tumor nodules, portal vein invasion by tumor, and large tumors identified as a poor-prognosis indicator.171–177 Large size of the primary tumor is a poor-prognosis indicator because of the increased frequency of vascular and lymphatic invasion by the tumor as well as growth along neighboring bile duct walls.178

Orthotopic liver transplantation has been described in patients with intrahepatic cholangiocarcinoma.177,179–181 The 1-year survival in series prior to 1990 was 29.4%, with only two of the patients undergoing liver transplantation being alive 5 years following the transplantation.181 Almost 90% of the patients who survived at least 90 days after the liver transplantation died of recurrent cholangiocarcinoma, frequently at extrahepatic sites. Recent small series of patients describe 5-year posttransplantation survival rates up to 53%.179,180 The improved survival is based on careful selection of cholangiocarcinoma patients for liver transplantation, specifically by not transplanting patients with lymph node metastases or invasion of major intrahepatic or extrahepatic blood vessels.

Hilar Bile Duct Cholangiocarcinoma

In 1890, Fardel first described a primary malignancy of the extrahepatic biliary tract.182 A report in 1957 described three patients with small adenocarcinomas involving the confluence of the left and right hepatic ducts.183 Such primary cholangiocarcinomas arising at the bifurcation of the extrahepatic biliary tree are known commonly as Klatskin tumors, following his report in 1965 of a larger series of patients with these lesions.184

Cholangiocarcinomas arising in the hilar bile ducts are relatively rare lesions. Extrahepatic biliary cancer has an incidence of 0.01% to 0.46% in autopsy series.184,185 Of 17,500 projected new cases of primary hepatobiliary cancers that occur annually in the United States, approximately 2,000 are Klatskin tumors.1

Prognostic Factors

In contrast to reports from two or three decades ago, most patients with hilar cholangiocarcinoma are now diagnosed before death. The most important factor affecting prognosis is the resectability of the tumor. Patients who undergo curative resection (margin-negative) have 3-year survival rates from 40% to 87% and 5-year survival rates between 10% and 73%.77,186–188 The wide range of survival rates is explained by variations in the incidence of factors that portend a poor prognosis in the various series. Significant determinants of improved prognosis in patients undergoing curative resection include well-differentiated tumors, absence of lymph node metastases, absence of direct tumor extension into the liver, papillary histology (versus nodular or sclerotic), serum bilirubin at presentation of less than 9 mg/dL, and a near-normal or normal performance status.77 Palliative resection, surgical bypass procedures, and various types of intubation and drainage procedures are associated with 3-year survival rates of 0 to 4%.186 Hilar cholangiocarcinomas have a poorer prognosis than do carcinomas arising in the middle or distal thirds of the extrahepatic bile duct, which is related directly to the presentation of hilar tumors at a more locally advanced stage with bilobar liver involvement by tumor and resultant lower rates of curative resection.189,190 However, like hilar cholangiocarcinoma, the presence of regional lymph node metastases reduces the 5-year overall survival rate following resection of middle or distal third bile duct cancer to 21% compared to the 65% survival rate in patients with node-negative disease.190

Pathologic features of the bile duct cancer are predictors of outcome. Prognosis is affected adversely if the tumor infiltrates the serosa of the bile duct, invades directly into the liver, demonstrates vascular invasion, or has metastasized to regional lymph nodes.191 Histologic type and grade also are important factors. Patients with the relatively unusual papillary bile duct adenocarcinoma have the most favorable prognosis, with 3-year survival rates up to 75%.186,191,192 Patients with the more common nodular or sclerotic types of hilar cholangiocarcinoma have 3-year survival rates of less than 30%. A pathologic study that correlated gross tumor type with patterns of spread provides evidence that may explain the observed differences in survival outcomes. Papillary and superficial nodular tumors spread predominantly by mucosal extension, rarely invading the deeper layers of the bile duct wall or lymphatic channels, whereas nodular infiltrating or diffuse infiltrating tumors spread by direct or lymphatic extension in the submucosa.193 The distance of mucosal or submucosal spread away from the gross tumor can be as great as 30 mm, but there were no local or anastomotic recurrences if at least a 5-mm tumor-free margin was attained. Patients with well or moderately differentiated carcinomas have a 3-year survival rate of up to 51%, whereas no patient with a poorly differentiated carcinoma survived longer than 2 years.191

Treatment

Resection

Resection of a hilar cholangiocarcinoma affords the patient the best chance for significant survival; however, 5-year survival rates after resection of hilar cancers are 40% in the most hopeful reports and 10% or less in other accounts. Long-term survival rates after resection of middle or distal common bile duct cholangiocarcinomas, the latter requiring pancreaticoduodenectomy, are generally higher compared with hilar tumors.99 This is most likely related to higher rates of margin-negative resection with middle or distal extrahepatic bile duct tumors and the absence of direct tumor extension into the liver.

The patterns of failure after curative extrahepatic bile duct resection for hilar cholangiocarcinoma have been described in a few series of patients (Table 102-3).194 Locoregional tumor recurrence developed in a high percentage of patients, with failure in the liver (62%), tumor bed (42%), and regional lymph nodes (20%). The caudate lobe is the most frequent site of liver recurrence. Regional lymph nodes include porta hepatis, retroduodenal, and perigastric node groups along the gastrohepatic ligament. Distant metastasis develops in the majority of patients who exhibit a locoregional recurrence; however, it was the site of first failure in only 24%.

Table 102-3. Sites of Tumor Recurrence after Curative Resection of Proximal Hilar Cholangiocarcinomas.

Table 102-3

Sites of Tumor Recurrence after Curative Resection of Proximal Hilar Cholangiocarcinomas.

Detailed anatomic studies have offered an explanation for the high incidence of liver and local recurrence following resection of a hilar cholangiocarcinoma. In a series of 25 patients undergoing surgery for hilar cholangiocarcinoma, direct invasion of hepatic parenchyma at the hilum was noted in 12 patients (46.2%), with 11 patients (42.3%) also having carcinoma extending into the bile ducts draining the caudate lobe or directly invading the caudate lobe parenchyma.195 A study of 106 adult human cadavers showed that 97.2% had bile ducts draining the caudate lobe that entered directly into the main left hepatic duct, right hepatic duct, or both.196 These caudate lobe bile ducts frequently enter the main left or right hepatic ducts within 1 cm of the proper hepatic duct. Thus, a carcinoma arising at the confluence of the right and left hepatic ducts need not be large to extend into the bile ducts draining the caudate lobe.

Because cholangiocarcinoma is known to spread along the wall of the bile ducts and because the caudate lobe and hepatic hilum are frequent sites of tumor recurrence following extrahepatic duct resection, a number of authors now recommend more aggressive resections to include the caudate lobe and hepatic hilar parenchyma.196–202 An understanding of the Bismuth-Corlette classification of hilar cholangiocarcinoma is useful in planning the extent and site of liver resection (Figure 102-11).203 The improved equipment and understanding of techniques requisite for a safe liver resection allow performance of aggressive extended resections, with little or no increase in operative morbidity and mortality. The median survival associated with a more radical surgical approach has varied from 10 to 37 months, with 5-year survival rates of 20% to 44% and 10-year survival rates as high as 14%.196–202 These studies clearly show that liver resection is worthwhile only if completely tumor-negative resection margins can be attained, because there were no 5-year survivors with positive resection margins. While aggressive surgical resection of hilar cholangiocarcinomas, including hepatic resection, provides the best chance for long-term survival, these operative procedures are associated with significant risk. The operative mortality rate in modern series ranges from 5% to 12%, with postoperative liver failure following an extensive liver resection being the most common cause of death.196–202 Surgical complications are reported in 25% to 45% of the surviving patients. Infectious complications are the most common postoperative problem, and preoperative placement of biliary stents with resultant contamination of the obstructed biliary tree increases the incidence of infection.204 Very extensive operations that include major hepatectomy, resection of the extrahepatic bile ducts, and en bloc pancreaticoduodenectomy have been used in patients with hilar cholangiocarcinoma.205 Given the operative mortality rate of at least 30% and a 100% rate of complications combined with rare survival for more than 2 years, such ultraradical procedures are of dubious value.

Figure 102-11. Bismuth-Corlette classification of hilar cholangiocarcinoma.

Figure 102-11

Bismuth-Corlette classification of hilar cholangiocarcinoma. Types 1 and 2 can be resected with excision of the extrahepatic bile duct with or without the hilar plate and caudate lobe. Types 3A and 3B can be resected with the addition of an en bloc right (more...)

A multimodal approach to reduce locoregional recurrence rates and improve survival after resection has been reported.206 Of 53 patients who underwent resection of a hilar cholangiocarcinoma, 38 received postoperative external beam radiotherapy to the resection bed at a dose of 50 to 60 Gy. In addition, 27 of these 38 patients received brachytherapy with 192 Ir seeds temporarily loaded into their transhepatic biliary stents in the area of the hepaticojejunostomies. These 27 patients received 20 Gy of internal radiation after completion of external beam radiotherapy. There was no significant difference in the 1-, 2-, and 3-year survival rates for patients who underwent resection, with or without radiotherapy, but there were no survivors past 3 years in the group without radiotherapy. The 5- and 10-year survival rates in the group receiving radiotherapy were 11% and 5%, respectively. Stented hepaticojejunostomies following resection of hilar cholangiocarcinomas allow access to the remaining biliary tree and can be used for diagnostic and therapeutic purposes.207

We recently reviewed our experience in patients with extrahepatic cholangiocarcinoma treated at the University of Texas MD Anderson Cancer Center.208 Of 91 patients evaluated between 1983 and 1996, 51 (56%) presented with unresectable disease and 40 (44%) underwent resection. The median survival for the resected patients was 22.2 months versus 10.7 months in patients with unresectable disease (p < .0001). Nine patients, 5 with hilar and 4 with distal common duct cholangiocarcinoma, were treated with preoperative chemoradiation therapy (continuous intravenous infusion of 5-fluorouracil at 300 mg/m2/day combined with external beam irradiation). Three of these 9 patients had a pathologic complete response to chemoradiation treatment; the remaining 6 patients had varying degrees of histologic response to treatment. The rate of margin-negative resection was 100% for the preoperative chemoradiation group compared to 54% for the group not receiving preoperative treatment (p < .01). The patients treated with preoperative chemoradiation had no operative or postoperative complications related to treatment. Thus, it appears that neoadjuvant chemoradiation for extrahepatic bile duct cancer can be performed safely, produces significant antitumor response, and may improve the ability to achieve tumor-free resection margins.

Liver transplantation

Total hepatectomy with immediate orthotopic liver transplantation (OLT) has been described in patients with hilar cholangiocarcinoma.177,179,180,209–222The 90-day mortality from hemorrhage, sepsis, and graft rejection was 23.1%. Of the patients who survived more than 3 months following transplantation, the median survival was 11 months in the series prior to 1992 but has improved to 23 months in the recent series. In the older series of patients, the 5-year survival rate was 5.0%. In patients who died later than 3 months after transplantation, death was due to tumor recurrence in 85.4%. The 5-year survival rate following OLT for extrahepatic cholangiocarcinoma in current studies is 25%, and in the highly selected subset of patients with stage I or II disease, 5-year survival is 73%.222 Nonetheless, due to the poor results in most reports, many transplantation centers no longer perform liver transplantations in patients with hilar cholangiocarcinoma. Liver transplantation for hilar cholangiocarcinoma should probably be considered only as part of a prospective protocol evaluating multimodality treatment.

Palliation

In general, curative surgical resection is possible in less than 30% of patients with hilar cholangiocarcinoma.77,186–188,223 In patients deemed unresectable on the basis of the findings of diagnostic studies, laparotomy can be avoided by placing percutaneous external drains or endoscopically placed endoprostheses.224,225 Conventional 10- or 12-French polyethylene endoprostheses have a high rate of occlusion and cholangitis.226 However, new expandable metal wall stents appear to have improved long-term patency rates and may be used to deliver palliative high-dose rate endoluminal brachytherapy.227,228 When unresectability is determined at the time of laparotomy, a decision must be made on a surgical bypass versus an operative intubation to provide drainage of the obstructed biliary tree. It is clear that techniques for surgical bypass, operative intubation, and percutaneous external drainage are equivalent in partial or complete relief of jaundice in 70% to 100% of patients.77 Seemingly, the only potential advantage to the patient who undergoes surgical bypass instead of operative intubation is the absence of an external drainage catheter in the former group.

The advantage of not having an external biliary drainage catheter or an internal endoscopically placed biliary endoprosthesis should not be underestimated because it is known that the incidence of cholangitis and occlusion or displacement of the catheter or endoprosthesis ranges from 28% to almost 100%.224–226,229 The assessments of quality of life in patients with hilar cholangiocarcinoma who undergo surgical bypass, operative intubation, or percutaneous drainage have not demonstrated a distinct advantage for any one type of palliative treatment.77,230 However, some studies suggest that the duration of well-being is longest in patients undergoing surgical bypass procedure.230–233 There is no significant difference in survival related to the type of palliative procedure employed to relieve biliary obstruction; the median survival in patients undergoing palliative drainage is 8 months or less. Nonetheless, effective palliation of biliary obstruction in patients with unresectable hilar cholangiocarcinomas is important because 50% will survive for at least 1 year, 20% will live for 2 years, and 10% will live 3 years or longer.234

Chemotherapy

Given the high percentage of unresectable hilar cholangiocarcinomas, various chemotherapeutic regimens and radiotherapeutic regimens have been used in the hope of providing improved palliation and prolongation of survival. Adjuvant radiotherapy also has been employed following resection of tumors to reduce locoregional recurrence and potentially to improve survival.206 Unfortunately, reports describing chemotherapy or radiotherapy for hilar cholangiocarcinoma rarely describe the treatment results in more than 10 to 20 patients.

A 1988 review of systemic chemotherapy for bile duct cancer noted that 97 patients had been treated with nine different treatment programs.235 Mitomycin-C, doxorubicin, and 5-fluorouracil are the agents that have shown the greatest activity against cholangiocarcinoma. The collective partial response rate in the 97 patients was 29%, with no complete responses. The median survival of these patients receiving systemic chemotherapy ranged between 6 and 11 months. No reports since 1988 have indicated a better response rate with systemic chemotherapy. With no significant increase in survival and considering the quality-of-life issues related to chemotherapeutic toxicity, systemic chemotherapy has not demonstrated a distinct advantage in patients with hilar cholangiocarcinoma.

Regional chemotherapy by hepatic artery infusion has some potential advantages over systemic chemotherapy. The proximal bile duct receives its arterial blood supply from the hepatic artery so that an increased concentration of drug can be delivered directly to the region of the tumor, including the liver. By using drugs such as 5-fluorouracil and floxuridine, systemic exposure to drug is limited because of the high rate of hepatic extraction of these agents. There are reports of 46 hilar cholangiocarcinoma patients who have been treated by hepatic artery chemotherapy infusion.235,236 The partial response rate for the entire group of patients was 43%, but there were no complete responses and no significant prolongation of survival. Hepatic artery infusion chemotherapy may have a role in the palliation of patients with hilar cholangiocarcinoma, but currently available drugs do not provide improved survival.

Radiotherapy

Radiotherapy for bile duct cancer is even more confusing due to the various types, doses, routes of administration, and association with resected and unresected tumors, all in small numbers of patients. Internal radiation with 192Ir wires or seeds may have a palliative role in improving the patency of obstructed bile ducts; however, the number and frequency of episodes of cholangitis were not reduced, so the overall benefit is uncertain.237 Internal radiation has been associated with prolongation of survival to an average of 16 months, and occasionally patients with unresectable disease survived more than 5 years.238–240 Although the use of external beam radiotherapy alone to treat patients with unresectable hilar cholangiocarcinoma has not provided significant differences in overall patient survival, rare long-term survivals have been reported.241 Intraoperative radiotherapy also has been evaluated in association with resectable and unresectable tumors.242,243 Again, there is a suggestion of a slight prolongation of survival in patients with unresectable tumors, but the most interesting use of intraoperative radiotherapy may be as an immediate surgical adjuvant in the resected high-risk tumor bed.

At the University of Texas MD Anderson Cancer Center, we have employed a treatment technique of four-field irradiation directed to the porta hepatis to a dose of 45 Gy. Patients receive 5-FU by continuous intravenous infusion during the 5-week course of external beam irradiation. Further radiotherapy is administered by a reduced treatment field using an arc rotation to deliver an additional dose of 15 to 20 Gy to the primary target volume.244 We have used this treatment plan for patients who have had a surgical biliary bypass procedure and do not have transcutaneous biliary drainage catheters. This radiation technique envelops the target and a small portion of the duodenum with a high dose. However, in patients with existing external biliary drainage tubes, a combination of external beam plus endoluminal boost irradiation is an attractive treatment program. The favored treatment sequence is to start with external beam irradiation to obtain tumor regression, which provides a better dose distribution from the endoluminal boost irradiation to treat any residual tumor. We currently use high-dose rate 192 Ir implants in a fractionated treatment schedule of 4 Gy daily for 5 consecutive days. The fractionated high-dose rate treatment, along with the favorable dose distribution obtained with endoluminal therapy, may decrease late-occurring duodenitis. For patients in whom there are technical or medical limitations for fractionated therapy, a single 20 Gy boost with the high-dose technique can be used.

The use of endoluminal 192Ir alone for palliative treatment of patients with unresectable hilar bile duct cancers has been reported.245 Endoluminal doses ranged from 15 Gy to 35 Gy when combined with external beam irradiation (usually 45B50 Gy), or when endoluminal doses of up to 60 Gy were used alone. The dose reference point may vary from 0.5 cm to 1.0 cm from the central catheter. The total nominal doses of external beam plus endoluminal boost irradiation are between 60 Gy and 70 Gy to the tumor, and although this range exceeds the liver and small intestine tolerance, the highest doses are confined to a small volume of tissue. The median survival for patients treated by this endoluminal method, with or without external beam irradiation, is 15 to 18 months. Several patients survived for more than 4 years after treatment; however, the majority of patients had local failure.

Complications with the endoluminal boost have been related mainly to cholangitis, which occurs, to some degree, in nearly all patients. Septic cholangitis and death rarely occur (in less than 15% of the patients) because antibiotic therapy is usually effective. Prophylactic antibiotics are not recommended because flora colonizing the catheter tract become drug resistant. A better approach to reduce sepsis is to remove the drainage catheter whenever there is cholangiographic evidence of bile duct patency following treatment.

Another specialized technique that has been used for boost treatments of biliary tract cancers is electron-beam intraoperative radiotherapy (EB-IORT).246 This technique improves the dose distribution by concentrating treatment directly on the target volume, while the uninvolved liver, adjacent duodenum, and stomach are excluded from the treatment field during surgery. Nine patients with advanced proximal biliary cancers were treated with EB-IORT in one report.247 Patients were considered for IORT if they had residual gross or microscopic disease confined to the region of the porta hepatis or if there was adjacent hepatic parenchymal involvement by tumor. The field size was 7 to 10 cm in diameter, and the IOR dose was 10 to 22 Gy. In five patients, additional external beam irradiation was given. An analysis of the patients treated with EB-IORT was made with concurrent groups who were treated without radiation or with external-beam irradiation with or without 192 Ir endoluminal boost irradiation. There was a significant improvement in median survival for the patients receiving high-dose irradiation, regardless of the boost technique, compared to those who were not irradiated (13 vs 4.6 months, respectively). Duodenitis occurred in most of the patients who received any form of irradiation, and portal vein thrombosis was observed in one patient who was treated with EB-IORT.

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2003, BC Decker Inc.
Bookshelf ID: NBK13176

Views

  • Cite this Page

Related Items in Bookshelf

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...