Cholangiocarcinoma: Molecular targeting strategies for chemoprevention and therapy


  • Alphonse E. Sirica

    Corresponding author
    1. Division of Cellular and Molecular Pathogenesis, Department of Pathology, Virginia Commonwealth University School of Medicine, Medical College of Virginia Campus, Richmond, VA
    • Division of Cellular and Molecular Pathogenesis, Department of Pathology, Virginia Commonwealth University School of Medicine, Medical College of Virginia Campus, P. O. Box 980297, Richmond, VA 23298-0297
    Search for more papers by this author
    • fax: 804-628-0375 or 804-828-9749

  • Conflict of interest: Nothing to report.


Cholangiocarcinomas are devastating cancers that are increasing in both their worldwide incidence and mortality rates. The challenges posed by these often lethal biliary tract cancers are daunting, with conventional treatment options being limited and the only hope for long-term survival being that of complete surgical resection of the tumor. Unfortunately, the vast majority of patients with cholangiocarcinoma typically seek treatment with advanced disease, and often these patients are deemed poor candidates for curative surgery. Moreover, conventional chemotherapy and radiation therapy have not been shown to be effective in prolonging long-term survival, and although photodynamic therapy combined with stenting has been reported to be effective as a palliative treatment, it is not curative. Thus, there is a real need to develop novel chemopreventive and adjuvant therapeutic strategies for cholangiocarcinoma based on exploiting select molecular targets that would impact in a significant way on clinical outcome. This review focuses on potential preventive targets in cholangiocarcinogenesis, such as inducible nitric oxide synthase, cyclooxygenase-2, and altered bile acid signaling pathways. In addition, molecular alterations related to dysregulation of cholangiocarcinoma cell growth and survival, aberrant gene expression, invasion and metastasis, and tumor microenvironment are described in the context of various clinical and pathological presentations. Moreover, an emphasis is placed on the importance of critical signaling pathways and postulated interactions, including those of ErbB-2, hepatocyte growth factor/Met, interleukin-6/glycoprotein130, cyclooxygenase-2, vascular endothelial growth factor, transforming growth factor-β, MUC1 and MUC4, β-catenin, telomerase, and Fas pathways as potential molecular therapeutic targets in cholangiocarcinoma. cholangiocarcinoma. (HEPATOLOGY 2005;41:5-15.)

Cholangiocarcinoma is the collective term used to describe cancers arising within the intrahepatic and extrahepatic biliary tract. More than 90% of cholangiocarcinomas are adenocarcinomas, with different histological variants being recognized, including adenocarcinoma, papillary adenocarcinoma, intestinal-type adenocarcinoma, and mucinous adenocarcinoma. Cholangiocarcinomas are classified further according to grade as being well, moderately, or poorly differentiated, with the classic diagnosis being well- to moderately-differentiated ductal adenocarcinoma. Desmoplastic reaction of variable degrees is also a common feature of cholangiocarcinoma. Although cholangiocarcinomas can occur anywhere within the biliary tract, approximately 40% to 70% occur at the liver hilum at or in close proximity to the bifurcation of the right and left hepatic ducts, whereas 5% to 20% of cholangiocarcinomas develop within liver. The term peripheral cholangiocarcinoma is used to describe those cancers originating within liver, typically as a solitary or multifocal mass, whereas cholangiocarcinomas originating at the liver hilum are known as hilar, perihilar, or Klaskin tumors.

Macroscopically, intrahepatic cholangiocarcinomas have been classified further as being either (1) a mass-forming type, (2) a periductal infiltrating type, or (3) an intraductal growth type, with the most commonly encountered types being the desmoplastic infiltrating nodular or diffusely infiltrating varieties (hilar cholangiocarcinomas are typically observed as being sclerosing ductal adenocarcinomas). The intraductal growth type is the least common and is characterized predominantly by intraductal growth with little or no extension beyond the bile duct walls. In comparison, cholangiocarcinomas with a highly infiltrative growth pattern exhibit lymphatic and intrahepatic metastases. Hilar invasive-type cholangiocarcinomas have been observed to exhibit perineural invasion and nodal involvement more frequently than do peripheral-type tumors. Intraductal papillary neoplasm of the liver exhibiting frequent gastroenteric metaplasia, overproduction of mucin, and progression to intrahepatic mucinous adenocarcinoma has also been recently characterized as forming a spectrum of biliary neoplasms occasionally associated with hepatolithiasis and that closely resemble intraductal papillary mucinous neoplasms of the pancreas.1

Currently, between 2,000 and 3,000 new cases of cholangiocarcinoma occur annually in the United States.2 Recent epidemiological studies have called attention to the apparent fact that this devastatingly lethal cancer has been increasing worldwide in both incidence and mortality over the past 2 to 3 decades,3–5 although the cause of these increases remains unclear. The challenges posed by cholangiocarcinoma continue to be daunting, in large part because of high morbidity and mortality rates, and the fact that early diagnosis of cholangiocarcinoma is difficult, with most patients having advanced incurable disease at presentation.

Clinically, treatment options for cholangiocarcinoma are limited, and conventional chemotherapy and radiation therapy to date have been notably ineffective in improving long-term survival. Presently, the only hope for long-term survival is complete resection of the tumor. Unfortunately, the vast majority of patients (i.e., 70%- 90%) with cholangiocarcinoma at presentation are not good candidates for curative surgery, and in those that have undergone complete surgical resection, the recurrence rate remains quite high.

Patients with unresectable cholangiocarcinoma have a survival that is generally less than 12 months after diagnosis. In comparison, the 5-year survival rates of patients in selected series who have undergone curative resection have ranged from 0% to approximately 40%,6, 7 with 5-year survival rates after resection for cholangiocarcinoma complicating primary sclerosing cholangitis (PSC) found to be less than 10%.7 Photodynamic therapy recently was shown to have a promising effect on improving palliation and survival in patients with unresectable hilar cholangiocarcinoma.8, 9 However, photodynamic therapy is not curative and further trials are needed to assess its effectiveness fully in combination with biliary stenting as a neoadjuvant therapy for cholangiocarcinoma. Another relatively recent and novel approach to palliation of unresectable cholangiocarcinoma involves the use of carboplatin-impregnated plastic biliary stents as a means of delivering cytotoxic chemotherapy directly to the tumor tissue.10 This approach is interesting, but it also is limited by the fact that it is only effective against local and not disseminated tumor. Last, liver transplantation is not generally indicated for cholangiocarcinoma, in large part because of the high tumor recurrence rates, although a small, selected group of patients with early-stage disease have benefited from this procedure.6, 7, 11

Based on the fact that cholangiocarcinoma remains for the most part a fatal disease with limited treatment options for prolonged survival, together with the fact that this lethal cancer is being encountered more frequently in clinical practice, there is now a real and urgent need to focus on developing novel chemopreventive and therapeutic strategies aimed at exploiting select molecular targets aberrantly expressed during cholangiocarcinogenesis that would impact in a significant way on clinical outcome.


PSC, primary sclerosing cholangitis; iNOS, inducible nitric oxide synthase in hepatic cell types; COX-2, cyclooxygenase 2; EGFR, epidermal growth factor receptor; TGF-β, transforming growth factor; RTKs, receptor tyrosine kinases; HGF, hepatocyte growth factor; hTERT, human telomerase reverse transcriptase; VEGF, vascular endothelial growth factor.

Chronic Inflammation, Bile Acids, and Chemopreventive Strategies

Cholangiocarcinomas generally are believed to arise as a result of malignant transformation of epithelial cells (cholangiocytes) lining the intrahepatic and extrahepatic bile ducts.2, 7 A hepatic stem cell origin of intrahepatic cholangiocarcinoma also has been proposed,12 but not as yet proven. However, the histogenesis of intestinal-type cholangiocarcinomas and of combined hepatocellular and cholangiocarcinoma observed in experimental rodent models of liver carcinogenesis13, 14 and in humans15, 16 is consistent with the concept that at least some subtypes of cholangiocarcinoma may be derived from a pluripotent liver stem cell.

Despite uncertainties as to the actual identity of the initiated cell type in cholangiocarcinogenesis, it is apparent that cholangiocarcinomas develop in the biliary tract as the result of a multistep process that seems to follow a sequence that includes early hyperplastic, and in some cases metaplastic, changes, followed by the appearance of biliary dysplastic lesions, and leading eventually to the development of cancer. In humans, cholangiocarcinomas occur most commonly between the 5th and 8th decade of life, suggesting a relatively long latency period to tumor development, although cholangiocarcinoma also occurs in persons younger than 50 years, with PSC being a prominent risk factor in this group of patients. In most cases of cholangiocarcinoma, the inciting cause is unknown, but chronic inflammation and cellular injury within bile ducts, together with partial obstruction of bile flow,2, 7 manifested in high-risk conditions for cholangiocarcinoma development, such as PSC, hepatolithiasis, and liver fluke infestation by Opisthorchis viverrini or Clonorchis sinensis, Caroli's disease, and congenital choledochal cysts,2, 7 clearly seem to be relevant predisposing factors in the pathogenesis of cholangiocarcinoma.

Figure 1 depicts salient cellular events associated with chronic inflammation and bile acids that have been postulated to contribute to the cholangiocarcinogenic process.7, 17 As depicted, excess generation of endogenous nitric oxide resulting from induction by proinflammatory cytokines of inducible nitric oxide synthase (iNOS) in the inflamed biliary tract has been linked to bile duct cell damage and potentially to cholangiocarcinogenesis18, 19 by virtue of its demonstrated effects on potentiating oxidative DNA damage and on mediating inhibition of DNA repair,20, 21 thereby favoring the possibility of oncogenic mutations. In addition, nitric oxide has been demonstrated to inhibit apoptosis downstream of cytochrome c by a mechanism involving nitrosylation of caspase-9.22 Moreover, nitric oxide was shown to cause bile ductular cholestasis by inhibiting cyclic-AMP–dependent ion transport functions,19 and may also favor cholangiocarcinogenesis by possibly contributing to tumor immunosuppression.23 Increasing evidence also suggests that cyclooxygenase-2 (COX-2), the inducible isoform of prostaglandin endoperoxide synthase (the rate-limiting enzyme in the biosynthesis of prostaglandins from arachidonic acid) is playing an important role in cholangiocarcinogenesis, both in experimental rodent models2, 24, 25 and in humans,2, 26, 27 likely through its actions on promoting cholangiocarcinoma cell growth and survival.27–29 The possibility of iNOS contributing to the upregulation of COX-2 in the biliary tract is supported by the recent findings of Ishimura et al.,30 who demonstrated that iNOS induces COX-2 expression in immortalized, but nonmalignant, murine cholangiocytes. In addition, oxysterols (oxygenated derivatives of cholesterol identified in bile from patients with inflammatory biliary diseases) recently were reported to induce COX-2 in cultured KMBC cholangiocarcinoma cells.31 Bile acids, including deoxycholic acid, a potent tumor promoter in the hamster model of cholangiocarcinogenesis,32 also were recently shown to induce COX-2 in cultured immortalized human H69 cholangiocyte and KMBC cell lines through a mechanism involving transactivatation of the epidermal growth factor receptor (EGFR).33, 34 Bile acids further have been shown to block protein degradation of myeloid cell leukemia protein 1, a potent antiapoptotic protein of the Bcl-2 family, via activation of EGFR signaling.35 It is not known, however, if bile acids or oxysterols directly can induce iNOS in cholangiocytes, but a possible link in situ between enhanced iNOS and COX-2 expression as being relevant to human cholangiocarcinogenesis is suggested by the results of immunohistochemical studies demonstrating strong immunoreactivity for iNOS19 and for COX-226 in cholangiocytes lining the large bile ducts of patients with advanced PSC. In contrast, in these same studies, major bile duct immunostaining for iNOS was reported to be mostly negative and for COX-2 to be notably weaker in patients with primary biliary cirrhosis, a nonrisk inflammatory condition for cholangiocarcinoma. This noted discrepancy in observed immunostaining reactions for iNOS and COX-2 in the PSC versus primary biliary cirrhosis livers could relate to differences in the severity of the inflammatory reaction and the degree of biliary obstruction produced in the major bile ducts in each of the two biliary disease conditions. Nevertheless, the higher levels of expression of iNOS and COX-2 detected in the large bile ducts of the PSC livers, in contrast to the primary biliary cirrhosis livers, support the possibility that iNOS and COX-2 are likely contributing in a coordinated manner to cholangiocarcinogenesis in high-risk conditions that are associated with longstanding inflammation within the biliary tract, together with impaired bile flow and altered bile acid composition.

Figure 1.

Action of inflammatory cytokines and bile acids in cholangiocarcinogenesis: potential targets for chemoprevention. “?” indicates not tested preclinically or clinically against cholangiocarcinogenesis. See Chronic Inflammation, Bile Acids, and Chemopreventive Strategies for details. TNFα, tumor necrosis factor α; IL-6, interleukin-6; iNOS, inducible nitric oxide synthase; NO, nitric oxide; COX-2, cyclooxygenase-2; EGFR, epidermal growth factor receptor; Mcl-1, myeloid cell leukemia protein 1; PGE2, prostaglandin E2.

Based on the scheme shown in Fig. 1, it seems reasonable to postulate that selective targeting of upregulated iNOS and COX-2, together with altering the composition of the hydrophobic bile acid pool to reduce tumor promoting bile acids like deoxycholic acid, may provide a rational strategy for the chemoprevention of cholangiocarcinoma in high-risk conditions such as PSC in a manner analogous to that which has been most notably demonstrated for colorectal cancer chemoprevention. In this context, the hydrophilic bile acid ursodeoxycholic acid, which is used in the treatment of PSC, has been reported to decrease the risk of colorectal dysplasia in patients with ulcerative colitis and PSC,36 although it is not yet known if ursodeoxycholic acid also reduces the risk of dysplasia in the biliary tract of patients with high-risk conditions for cholangiocarcinoma. Selective iNOS inhibitors also have been shown to exhibit chemopreventive effects in rodent models of colorectal and esophageal carcinogenesis,37, 38 but to date these agents have not been tested experimentally in animal models of cholangiocarcinogenesis that recapitulate salient features of the human disease, such as iNOS overexpression and COX-2 upregulation. Numerous studies further have demonstrated select COX-2 inhibitors to be chemopreventive for colorectal cancer development in both experimental animal models and in humans,39 but like iNOS inhibitors, COX-2 inhibitors have not been assessed for their chemopreventive effects in preclinical in vivo models of cholangiocarcinogenesis. Moreover, the chemopreventive potential of combinational treatments with agents such as ursodeoxycholic acid or COX-2 inhibitors with select iNOS inhibitors also needs to be evaluated experimentally to realize the full potential and safety of such chemopreventive strategies for cholangiocarcinogenesis. Here it may be noteworthy that combined administration of the iNOS inhibitor SC-51 together with the COX-2 inhibitor celecoxib was found to be more effective than either agent alone in suppressing colonic aberrant crypt foci formation in the azoxymethane rat model of colon carcinogenesis.37 It is also important to determine the limitations of such chemopreventive strategies. For example, treatments with such agents as select COX-2 inhibitors may be contraindicated as chemopreventive treatments for PSC patients by virtue of their potential to induce hepatic toxicity,40, 41 as well as other potential toxic effects. Nevertheless, carefully planned studies are now needed to validate the chemopreventive targets depicted in Fig. 1, as well as to assess the safety and translational relevance of such treatments for patients with high-risk conditions for cholangiocarcinogenesis.

Molecular Alterations and Variations

Significant progress has been made over the past decade in defining molecular alterations associated with cholangiocarcinoma. Table 1 summarizes the most studied molecular alterations that have been described to date for human cholangiocarcinomas in relation to dysregulation of cell growth and survival pathways, aberrant gene expression, invasion and metastasis, and tumor microenvironment. It is beyond the scope of this review to provide a detailed analysis of each of these factors and their relationship to the molecular pathogenesis of cholangiocarcinoma. However, the reader is referred to the several recent reviews7, 42–47 for an updated analysis of the molecular pathological features of cholangiocarcinoma. This section instead highlights such factors as tumor location, subtype, and grade, which have been found to be associated with variations in the frequencies of expression of selected molecular alterations and which need to be considered when assessing outcome and in devising molecular therapeutic strategies for cholangiocarcinoma.

Table 1. Molecular Alterations in Cholangiocarcinoma
  1. Abbreviations: pt. mut., point mutation; LOH, loss of heterozygosity; methyl., methylation; prom. methyl., promoter methylation; allel., allelic; IL-6, interleukin-6; IL-6R, interleukin-6 receptor; gp130, glycoprotein 130; bFGF, basic fibroblast growth factor; TGF-β1, transforming growth factor β1.

Autonomous growth signaling
 HGF/Met overexpression and activation
 IL-6, IL-6R/gp130 overexpression and activation
 ErbB-2 overexpression and activation
 K-ras activating mutations
 BRAF activating mutations
 COX-2 upregulation
Abnormalities of DNA mismatch repair
 Microsatellite instability
 Telomerase ↑
Inactivation of tumor suppressor genes
 p53: by pt. mut.; LOH; mdm2
 APC: by LOH; methyl.
 p16INK4a: by pt. mut.; prom. methyl.; allel. loss
 DPC4/Smad4: by pt. mut.; allel. loss
Cell cycle dysregulation
 Cyclin D1 ↑, p53 ↑, pRb ↓, p16INK4a ↓, p21waf1/cip1 ↓, p27kip1 ↓, p57kip2
Aberrant mucin antigen expression and homeodomain gene expression
 Bcl-2 ↑, Bcl-XL ↑, Mcl-1 ↑, COX-2 ↑, Fas ↓
Angiogenesis and desmoplasia
 Expression of VEGF, COX-2, bFGF, HGF, TGF-β1
Invasion and metastasis
 E-Cadherin ↓
 α-Catenin ↓
 β-Catenin ↓
 Human aspartyl (asparaginyl) β-hydroxylase ↑
 WISP1v ↑

Wide variation in the reported range of frequencies (i.e., see Rashid43) of many of the abnormalities listed in Table 1 make it unlikely that a single pathway to cholangiocarcinoma development based on a common set of molecular and genetic defects will be established. Nevertheless, the current data make it feasible to attempt separating some of these alterations in terms of specific morphological, pathological, and clinical presentations of cholangiocarcinoma. K-ras mutations, typically at codon 12, have been reported to be less frequently detected in peripheral cholangiocarcinomas than in hilar cholangiocarcinomas48, 49 and seem to be occurring at an increasingly higher incidence in bile duct cancers arising distally in the common bile duct.48, 50 Intrahepatic cholangiocarcinomas of the mass-forming type were further found to harbor K-ras mutations less frequently than those of the periductal infiltrating type,51, 52 with the incidence of K-ras mutations having been reported to also be higher in cholangiocarcinoma patients with lymph node metastasis than in those without lymph node metastasis.49 Potential activating mutations of the BRAF gene, one of the human isoforms of Raf that is activated by oncogenic Ras, also have been detected in cholangiocarcinoma.53 However, in this limited study, neither the mutational status of BRAF, K-ras, or both seemed to influence the survival of the patients with cholangiocarcinoma. Mutational inactivation of the putative tumor suppressor gene DCP4/Smad4, an important downstream component of the transforming growth factor β (TGF-β) signaling pathway, also has been found to occur more commonly in distal bile duct cancers than in more proximal bile duct and intrahepatic tumors.54, 55

Alterations in p537, 47 and p16INK4a7, 56 are frequently detected in cholangiocarcinomas and are likely contributing to oncogenesis in the biliary tract. Frequency of overexpression of p53 was reported to be significantly higher in distal compared with proximal bile duct cancers,55 as well as in nonpapillary intrahepatic cholangiocarcinomas than in the intraductal papillary subtype.45p53 dysregulation further has been postulated to represent a mid-to-later genetic event in the development and progression of intrahepatic cholangiocarcinoma.45 In contrast, point mutations in the promoter region of p16INK4a seem to represent an apparent early event in PSC-associated cholangiocarcinogenesis.56 Other alterations that seem to occur early in cholangiocarcinogenesis include overexpression of the receptor tyrosine kinases (RTKs) ErbB-2 and Met2, 27, 57–59 and the upregulation of COX-2.2, 26, 27 Moreover, strong expression of ErbB-2,2, 27, 57 Met,27, 58, 59 COX-2,2, 27 Bcl-2,60 and Fas ligand61 have been detected more frequently in well-differentiated cholangiocarcinomas compared with moderate to poorly differentiated tumors, whereas downregulation of β-catenin and E-cadherin,62, 63 overexpression of aspartyl (asparaginyl) β-hydroxylase,64 and decreased expression of such factors as Fas receptor61 and p57kip265 correlated with higher histological grades of tumor. No correlation was found between p27kip1 expression and tumor grade, but low p27kip1 expression was shown to be an independent predictor of survival for patients with either peripheral or hilar cholangiocarcinoma.66 Cyclin D1 overexpression was more frequently observed in cases of intrahepatic cholangiocarcinoma with poor or moderate differentiation and with lymph node metastasis.67 Likewise, WISP1v expression was found to be associated with lymphatic and perineural spread of cholangiocarcinoma and poor clinical outcome.68 Human telomerase reverse transcriptase (hTERT) has also been detected in a high percentage of analyzed cases of intrahepatic cholangiocarcinoma, irrespective of tumor grade and subtype, as well as heterogeneously in dysplastic lesions, suggesting that acquired hTERT activity may reflect an early stage leading to cholangiocarcinoma development.69

MUC1 apomucin is frequently expressed in various subtypes of intrahepatic cholangiocarcinoma, including mass-forming and periductal infiltrating forms.70 High expression and cytoplasmic location of MUC1 correlated with lower survival rates for patients with mass-forming intrahepatic cholangiocarcinoma.71, 72 MUC4, which acts as a ligand for ErbB-2, was also recently demonstrated to be an independent risk factor for poor prognosis in patients with the mass-forming type of intrahepatic cholangiocarcinoma.73 Of further note is the finding that patients in this series with cholangiocarcinomas double positive for MUC4 and ErbB-2 at presentation had the worst outcome. Serum levels of MUC5AC, another mucin frequently expressed in intrahepatic cholangiocarcinomas, have also been reported to correlate with tumor burden and poor patient survival.74 In contrast, MUC2, an intestinal-type mucin that is selectively expressed predominantly in well-differentiated and noninvasive mucinous-type intraductal cholangiocarcinomas with gastrointestinal differentiation,75 predicts a more favorable prognosis.44, 71 MUC2 expression in intraductal papillary neoplasms of the liver was recently demonstrated to be closely related to aberrant expression of the caudal-related homeodomain intestine-specific transcription factor CDX2.75 CDX1, another member of this homeodomain intestine-specific transcription factor family, also has been demonstrated to be associated with mucin-producing glands in intestinal-type intrahepatic cholangiocarcinomas induced in the liver of furan-treated rats.76 MUC expression profiles, therefore, seem to have predictive value for patient outcome and may be useful in tailoring specific treatment strategies for cholangiocarcinoma subtypes, such as those with or without cellular and genetic evidence of gastrointestinal differentiation.

RTKs, Interactive Signaling Pathways, and Therapeutic Targets

Figure 2 represents a simplified, but useful, scheme depicting aberrant interactive molecular pathways having promise as potential selective therapeutic targets for suppressing tumor growth, inducing apoptosis, and blocking malignant cell invasion and inhibiting tumor angiogenesis in primarily well to moderately differentiated cholangiocarcinomas. The solid arrows represent established interactions based directly on data derived from analyses of human and rodent biliary tract cancers and cholangiocarcinoma cells,24, 27, 77–82 whereas the dashed arrows reflect postulated interactions extrapolated from recent findings reported for other malignant tumor types, including breast, prostate, colorectal, and pancreatic carcinoma cells. Briefly, various cholangiocarcinoma cell lines and tumors have been demonstrated to profoundly overexpress constitutively activated ErbB-2,2, 24, 27, 42, 78 to exhibit evidence of autocrine signaling loops for hepatocyte growth factor (HGF)/Met77, 79 and interleukin-6/interleukin-6 receptor/glycoprotein 13079 and to show bile acid–induced activation of EGFR.33, 34 Each of these RTK signaling pathways has been linked to COX-2 induction,24, 27, 33, 42 resulting in prostaglandin E2 overproduction.25, 78, 80 Prostaglandin E2, in turn, has been shown to be capable of transactivating EGFR83 and Met84 of activating the interleukin-6/glycoprotein 130 signaling pathway85 and of enhancing ErbB-286 and vascular endothelial growth factor (VEGF) expression87, 88 in cancer cells. It is of further interest that interleukin-6 as well as HGF have been recently demonstrated to stimulate arachidonic acid release and prostaglandin E2 production in human cholangiocarcinoma cell lines expressing COX-2 by a mechanism mediated by phosphorylation of cytosolic phospholipase A2.80 Moreover, prostaglandin E2 generation by COX-2 has been shown to inhibit Fas-mediated apoptosis in human cholangiocarcinoma cells by a mechanism apparently involving myeloid cell leukemia protein 1 upregulation.81 A novel mechanism of cell survival in which Fas becomes sequestered after its interaction with the HGF receptor Met, thereby preventing Fas self-aggregation and Fas ligand binding, has also been described for hepatocytes,89 which may also be significant for the suppression of cell death by apoptosis in cholangiocarcinoma.

Figure 2.

Postulated interactive molecular pathways as potential therapeutic targets in primarily well to moderately differentiated cholangiocarcinoma. See RTKs, Interactive Signaling Pathways, and Therapeutic Targets for details. MUC4, mucin 4; MUC1, mucin 1; IL-6R, interleukin-6 receptor; gp130, glycoprotein 130; EGFR, epidermal growth factor receptor; VEGFRs, vascular endothelial growth factor receptors; COX-2, cyclooxygenase-2; HGF, hepatocyte growth factor; VEGFs, vascular endothelial growth factors; PGE2, prostaglandin E2; TGF-β1, transforming growth factor β1.

In addition to playing a role in regulating the COX-2 pathway,86, 90 ErbB-2 has been reported to form a complex with the glycoprotein 130 subunit of the interleukin-6 receptor, suggesting an intimate link between ErbB-2 and interleukin-6 receptor signaling.91, 92 Proinflammatory cytokines like interleukin-6 were also observed to enhance transcription of COX-293, 94 and MUC1.95 ErbB-2 signaling has been further shown to regulate the expression of MUC196 and VEGF97, 98 and to facilitate induction of hTERT gene transcription and telomerase activity99 in cancer cells. Furthermore, ErbB-2 has been reported to interact with both MUC1100 and β-catenin.100, 101 MUC4, which triggers tyrosine phosphorylation of ErbB-2, was also recently shown to be upregulated by bile acids by a mechanism involving activation of the phosphatidylinositol 3-kinase signaling pathway.102

Like ErbB-2, MUC1100, 103 and Met104 have been demonstrated to interact with β-catenin, potentially contributing to a more invasive and metastatic phenotype. Interestingly, HGF has been demonstrated to induce Wnt-independent nuclear translocation of β-catenin, based on the association of activated Met with β-catenin,105 thereby leading to the likelihood of altered gene expression relevant to oncogenesis.

β-catenin106 and TGF-β182 have been demonstrated to regulate VEGF expression as well, and it has been suggested that overexpression of TGF-β1 and VEGF may contribute to the “angiogenic switch” and malignant phenotype in human cholangiocarcinoma.82 It also seems noteworthy that human cholangiocarcinoma cell lines insensitive to TGF-β1 when xenografted into nude mice yielded tumors that had a more prominent desmoplasia than tumors developed from TGF-β1–sensitive cell lines.107

Preclinical Results

Table 2 (108–120) lists the results of several recent preclinical studies supporting the therapeutic potential of selected targeting strategies against cancers of the biliary tract, including those aimed at many of the molecular targets depicted in Fig. 2. Recent results obtained with other gastrointestinal tumor cell types also support targeting of MUC4,121 hTERT,122 β-catenin,123 and the VEGF/VEGF-R124 and TGF-β signaling pathways125 as promising molecular therapeutic approaches to the treatment of cholangiocarcinoma. Combinational targeting of select RTK pathways (i.e., ErbB-2/EGFR) together with other targets (i.e., COX-2) shown in Fig. 2 have the potential of producing even more profound and potentially synergistic or cooperative antitumor and antiangiogenic effects for cancers overexpressing these gene products than single-agent therapies.78, 126, 127 Molecular and immunochemical marker expression profiling of biopsied or surgically resected tumor samples will be useful in identifying already cited as well as new selected targets for novel combinational therapeutics. COX-2, select RTK inhibitors, or both should also be considered for their ability to function as radiosensitizers, as well as to potentiate the effects of conventional cancer chemotherapeutic agents, such as gemcitabine, and potentially to enhance photodynamic therapy-mediated tumor responses. However, again it is balancing therapeutic effectiveness of such novel molecular therapeutic approaches with protection against potential host toxicity reactions that will be the challenge.

Table 2. Preclinical Approaches Supporting Potential Therapeutic Usefulness of Molecular Targeting Strategies Against Biliary Tract Cancer Cells
AgentClassBiliary Cancer Cell LineSpeciesExperimental ConditionBiological EffectsReferences
  1. Abbreviations: IL-6R, interleukin-6 receptor; PPAR γ, peroxisome proliferator-activated receptor γ; SCID, severe combined immunodeficiency.

Anti-IL-6RNeutralizing antibodiesKMCH-1HumanCell cultureAttenuated cell growthPark et al.108
HGF/NK4HGF antagonistGB-d1, HuCCT1HumanCell culture/in vivoInhibited HGF-induced invasion of GB-d1 and HuCCT1 cells in vitro; suppressed growth and invasion of GB-d1 tumor cell xenograftsDate et al.109
CelecoxibCOX-2 inhibitorC611B; HuCCT1, QBC939, SG231, CCLP1Rat; humanCell culture/in vivoInduced dose-dependent growth inhibition and apoptosis in vitro of rat and human cell lines; partially suppressed tumorous growth of rat C611B cells in vivoZhang et al.,25 Hayashi et al.,26 Wu et al.,28 Han et al.,29 Wu et al.110
COX-2 antisense vectorInhibitor of human COX-2 gene expressionQBC939HumanCell cultureDecreased in vitro cell proliferative indexWu et al.111
Emodin; GW572016ErbB-1/ErbB-2 receptor tyrosine kinase inhibitorsC611BRatCell cultureInduced dose-dependent growth inhibition and apoptosis in vitroLai et al.,78 Lai et al.112
Emodin combined with celecoxibPredominantly ErbB-2 receptor tyrosine kinase inhibitor and COX-2 inhibitorC611BRatCell cultureActed synergistically to enhance inhibition of anchorage-dependent and anchorage-independent cell growth and to increase apoptosis in vitroLai et al.78
15-deoxy-Δ12,14-prostaglandin J2PPARγ ligandHuCCT1, SG231, CC-LP-1, RBE, ETK-1HumanCell cultureInduced dose-dependent inhibition of cell growth of HuCCT1, SG231, and CC-LP-1 cells and apoptosis in RBE and ETK-1 cellsOkano et al.,113 Han et al.114
Antisense oligonucleotides against aspartyl (asparaginyl) β-hydroxylase (AAH)AAH gene expressionSSP-25, NEC, ETK-1, RBE, H-1HumanCell cultureSuppressed cholangiocarcinoma cell migration in vitro as a function of inhibition of AAH expressionMaeda et al.64
Lymphokine-activated killer cells sensitized to bispecific antibodies MUC1 × CD3 and MUC1 × CD28Adoptive immunotherapy against MUC1-positive target tumor cellsTFK-1HumanIn vivoInhibited xenografted tumor growth in SCID miceKayatose et al.115
Interferon-γPleiotropic cytokineSk-ChA-1HumanCell culture/in vivoEnhanced Fas-mediated apoptosis in cultured cholangiocarcinoma cells and inhibited tumorigenicity of the Fas-low subpopulation of Sk-ChA-1 cellsAhn et al.116
Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)/Apo2LTumor necrosis factor family cytokineTFK-1, HuCCT1HumanCell culture/in vivoInduced significant apoptosis in vitro and substantially inhibited tumorigenicity of HuCCT1 cells subcutaneously implanted into nude miceTanaka et al.117
Adenovirus vector expressing p27kip1Universal cyclin-dependent kinase inhibitorTFK-1, HuCCT1HumanCell culturep27kip1 overexpression induced apoptosis in cholangiocarcinoma cells via utilization of the Fas pathwayYamamoto et al.118
RGD-modified COX-2 promoter-driven herpes simplex virus thymidine kinase (HSV-TK) adenoviral expression vectorsToxin gene/prodrug paradigm of (HSV-TK)/gancyclovirOz, Sk-ChA-1HumanCell cultureElicited a cytocidal response in vitro in infected cholangiocarcinoma cells on gancyclovir administrationNagi et al.119
TauroursodeoxycholateBile acid conjugateMz-ChA-1HumanCell cultureInhibited cholangiocarcinoma cell growth in vitro via Ca2+-, PKC-α, and MAP kinase-dependent pathwaysAlpini et al.120

Although encouraging, results of the preclinical therapeutic studies shown in Table 2 in large part were obtained from in vitro experiments and from a limited number of in vivo studies involving tumor cell xenografts. However, to advance fully and to predict more realistically the potential clinical value of novel therapies against cholangiocarcinoma that are based on selective molecular targeting, such as targeting with select neutralizing antibodies, small drug RTK inhibitors, antisense vectors, and small interfering RNA therapeutics, there is a real need to establish preclinical testing strategies using animal models of cholangiocarcinoma closely resembling the human disease. In this context, a number of rodent models of cholangiocarcinogenesis currently exist, including Syrian hamsters exposed to a subcarcinogenic dose of dimethylnitrosamine and infected with Opisthorchis viverrini, as well as the furan-induced13 and thioacetamide-induced128 rat models of cholangiocarcinogenesis. Gallbladder adenocarcinomas also have been shown to develop at a 100% incidence in a transgenic mouse model overexpressing wild-type ErbB-2 under the control of the bovine keratin 5 promoter.24 Cholangiocarcinomas derived from the furan rat model further have been found to recapitulate many of the findings described for human cholangiocarcinomas, including c-erbB-2 and c-met overexpression and COX-2 upregulation.2, 25, 42, 77, 78c-met and c-erbB-2 also were recently reported to be overexpressed in thioacetamide-induced rat cholangiocarcinomas128 and ErbB-2–overexpressing gallbladder adenocarcinomas from bovine keratin 5. ErbB-2 transgenic mice exhibited constitutive upregulation of COX-2.24 Thus, it seems that such animal models may be well suited for use as preclinical platforms for testing molecular therapeutic and chemoprevention strategies that have the potential of translating to the human disease. The development of novel orthotopic cell transplantation models using established animal and human cholangiocarcinoma cell lines, such as the rat C611B cholangiocarcinoma cell line42 transplanted into livers of syngeneic Fischer 344 rats, also are likely to benefit research aimed at the experimental therapeutics of cholangiocarcinoma.


Cholangiocarcinoma continues to be a challenging cancer that requires innovative approaches to early diagnosis, prevention in high-risk populations, and in particular, the development of novel adjuvant therapies having potential for greatly improving long-term survival rates. There is no doubt, however, that with a better understanding of the complex interactive roles played by cancerous cholangiocytes and their microenvironment in affecting cholangiocarcinoma cell growth, invasion, and angiogenesis signaling pathways, powerful new molecular therapies aimed at critical malignant cholangiocyte and tumor stromal cell targets will be forthcoming.


The author thanks Ms. Deanna J. Ward and Ms. Jennifer L. DeWitt for assisting in preparation of the figures and in typing portions of the manuscript.