Review Genes and Cholangiocarcinoma

REVIEW

GENES AND CHOLANGIOCARCINOMA

Kanlayanee Sawanyawisuth

Department of Biochemistry, Faculty of Medicine, Khon Kaen University, Khon Kaen, Thailand

Abstract. Cholangiocarcinoma (CCA) is a major public health problem in the northeastern part of Thailand where the prevalence of this cancer is highest in the world. It is a slow growing tumor but highly metastatic with a poor prognosis. The traditional habit of eating raw fish, repeated exposure to liver fluke and consumption of nitrosamine-contaminated food are major risk factors. Understanding of carcinogenesis and metastasis of cholangiocarcinoma at the molecular level will provide tools for better prevention, diagnosis and treatment. This review focuses on CCA- associated genes and their possible applications.

INTRODUCTION furcation of the right and left hepatic ducts and the tumor mass is usually located within The term “cholangiocarcinoma” is used the liver (Uttaravichien et al, 1999). Central for primary tumor of the bile duct, includ- type refers to CCA that originates from in- ing intrahepatic, perihilar, and distal extra- trahepatic portion of the right or left hepatic hepatic tumors (Nakeeb et al, 1996). In Thai- duct (lobar duct) and the tumor is located land, cholangiocarcinoma (CCA) refers to mainly at the hepatic hilus, usually derived malignant tumor of the biliary tree includ- from the right or left hepatic duct or its di- ing the intrahepatic and extrahepatic por- rect branches or in the vicinity of hepatic tions, but not carcinoma of the gall bladder duct confluence. and the ampulla of Vater (Uttaravichien et al, 1999). EPIDEMIOLOGY Intrahepatic CCA, originating from intrahepatic biliary tree including right and left In humans, CCA is a relatively rare can- hepatic ducts, is classified according to the cer that almost always presents with an ex- main location along the biliary tree as pe- tremely poor prognosis (Carriaga and ripheral and central types. Peripheral intra- Henson, 1995; Farley et al, 1995). Even in hepatic CCA is defined as carcinoma of the those that have undergone complete surgi- intrahepatic bile duct peripheral to the bi- cal resection, the recurrence rate remains quite high and consequently the 5-year sur- Correspondence: Dr Kanlayanee Sawanyawisuth, vival rate ranges only from 0-40% (Gores, Department of Biochemistry, Faculty of Medi- 2003; Anderson et al, 2004). Apart from cer- cine, Khon Kaen University, Khon Kaen 40002, tain regions of Asia, in particular Northeast Thailand. Thailand where CCA is endemic, worldwide Tel: 66 (0) 4334 8386; Fax: 66 (0) 4334 8386 frequencies of this malignant neoplasm have E-mail: [email protected] generally been reported to range from 5%

Vol 40 No. 4 July 2009 701 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH to 30% of all liver cancers (Parkin et al, 1993). (Ohshima et al, 1994; Satarug et al, 1996). Although CCA is relatively infrequent compared to hepatoma, attention to CCA is now CHOLANGIOCARCINOGENESIS growing because both incidence and mortality rate of CCA are increasing in the Several mechanisms by which OV in- United States, the United Kingdom, Austra- fection may enhance the genesis of lia and overall worldwide (Patel, 2001; Khan cholangiocarcinoma have been proposed et al, 2002). The highest incidence rate, 188 (Sripa and Pairojkul, 2008). The primary per 100,000 of CCA, is in Khon Kaen, Thai- pathological change, ie epithelial desquama- land (Sriamporn et al, 2004). tion, may be due to mechanical irritation caused by the liver fluke and/or its excretory/secretory (ES) products. However, im- RISK FACTORS munopathological processes may also con- Primary sclerosing cholangitis (PSC) is tribute to the long-standing hepatobiliary a well-known risk factor for CCA in Western damage. Following liver fluke infection, in- countries (Rosen et al, 1991) and associates flammation, periductal fibrosis and proli- with ulcerative colitis. The lifetime risk of ferative body responses, including epithelial CCA among ulcerative colitis patients is 10- hyperplasia, goblet cell metaplasia, 15% (Lee and Kaplan, 1995). In addition, adenomatous hyperplasia, may result in pre- thorotrast used as a radiologic contrast agent disposing lesions that enhance a greater sus- in x-ray diagnosis was reported to cause CCA ceptibility of DNA to carcinogens. several decades after exposure (Liu et al, In Northeast Thailand, several N-nitroso 2002). compounds and their precursors exist in low In Asian countries, liver fluke infection levels in fermented food, such as fermented is the major risk for CCA. Clonorchis sinensis, fish (“pla-ra”), which is a daily northeastern a liver fluke endemically infested in China, diet (Sripa et al, 2007). These compounds are Korea, Japan and Taiwan, has been proposed possible primary carcinogens leading to CCA to be a propable cause of CCA (Hou, 1956). in this region. Apart from exogenous carcino- In Northeast Thailand, the traditional habit gens, endogenous nitrosation caused by the of eating raw freshwater and salt-fermented liver fluke infection has been extensively fish on a daily basis results in a local popu- studied both in animals and humans lation repeatedly exposed to liver fluke (Pairojkul et al, 1991; Srivatanakul et al, 1991; (Opisthorchis viverrini, OV) infection and ni- Bartsch et al, 1992; Thamavit et al, 1994). Sev- trosamine-contaminated food from early in eral human studies indicated that infected life. Epidemiological studies have revealed individuals have a higher endogenous a correlation between high prevalence of nitrosation level than uninfected individual CCA and liver fluke infection in this region (Srivatanakul et al, 1991). OV infection in ham- (Khan et al, 2008). In addition, experimental sters can induce nitric oxide synthase produc- animal models have demonstrated that di- tion in macrophages, mast cells and eosino- etary contamination of nitrosamines and phils in the inflamed areas surrounding the opisthorchiasis are strong predisposing fac- bile ducts. Recently, biomarkers of DNA tors for the genesis of CCA (Thamavit et al, damage, 8-oxo-7, 8-dihydro-2’-deoxyguanine 1993; Mitacek et al, 1999). Apart from exog- (8-oxodG) and 8-nitroguanine were demon- enous nitrosamines, endogenous nitro- strated in inflammatory cells and bile duct samines were also shown as risk factors epithelium of OV infected hamsters (Pinlaor

702 Vol 40 No. 4 July 2009 GENES AND CHOLANGIOCARCINOMA et al, 2003). Moreover, repeated infection with Lymphatic permeation in CCA is more com- OV mediates oxidative and nitrosative DNA mon than vascular invasion. Spreading damages sooner than a single infection along nerves is encountered particularly in (Pinlaor et al, 2004). tumor at the large portal tracts. Intrahepatic Both exogenous and endogenous nitro- metastasis develops in nearly all cases at samine formation may lead to DNA alkyla- relatively advanced stage. The incidence of tion and DNA base deamination in predis- regional lymph node metastasis is high in posed and inflamed tissues. Excess nitric CCA. Blood-borne spread occurs later, par- oxide produced by the infection could also ticularly to the lungs, but other sites include exert direct cytotoxic and mutagenic effects bone, adrenals, kidney, spleen and brain and increase cell proliferation. DNA dam- (Sripa and Pairojkul, 2000). aged biliary epithelium may then be trans- formed to malignant CCA. PROGNOSIS A recent study has demonstrated the Prognosis of intrahepatic CCA is gen- effect of OV ES product(s) in inducting NIH- erally unfavorable because loco-regional 3T3 mouse fibroblast cell proliferation in extension is usually advanced at the time of vitro by stimulating the expression of phos- diagnosis. Even after a resection, the out- phorylated retinoblastoma (pRB) and cyclin come for patients with advanced stage is D1, key proteins driving cells through the extremely poor (Khuntikeo, et al, 2008). Sev- G1/S transition point of the cell cycle eral prognostic factors for patients with re- (Thuwajit et al, 2004). Furthermore, gene section have been reported. The presence expression profile obtained from cDNA ar- of vascular invasion, lymphatic invasion, ray analysis showed significant changes of and lymph node metastasis is significant gene expression in several functional catego- poor prognostic predictor for intrahepatic ries including genes related to cell prolifera- CCA (Hanazaki et al, 2002). In addition, tion (Thuwajit et al, 2006). TGF-beta and EGF periductal infiltration, perineural invasion, signal transduction pathways were indi- portal vein invasion, intrahepatic metasta- cated as possible pathways of OV-driven cell sis, and two or more lymph node metastases proliferation (Thuwajit et al, 2006). predict shorter survival after surgery (Uenishi et al, 2001; Hirohashi et al, 2002; TUMOR SPREAD AND METASTASIS Suzuki et al, 2002) . IN CCA Several studies also showed a significant Various routes of tumor spread are seen relationship of survival rate to vascular per- in CCA. The recognized patterns include di- meation, extrahepatic metastases, and lym- rect invasion, infiltration along the biliary phatic, neural, and nodal involvement. Pa- tree, vascular and lymphatic permeation, tients with lymphatic, neural, or nodal in- and perineural and intraneural invasion volvement die early after surgery (Namieno (Sripa and Pairojkul, 2000). Direct invasion et al, 2001). Among the different routes of to the adjacent liver parenchyma through the metastasis, the presence of lymph node me- sinusoid is common in CCA. This may oc- tastasis has been reported in most studies to cur with or without inflammatory cell infil- be the worst prognostic factor after a resec- tration. On the other hand, infiltration along tion (Shirabe et al, 2002). The survival rate for the biliary tree usually shows periductal in- intrahepatic CCA patients positive for lymph filtration and variable desmoplastic reaction. node metastasis is lower than those negative

Vol 40 No. 4 July 2009 703 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH for lymph node metastasis (Taoka and expression in cancer tissues by immunohis- Kawarada, 1997; Okabayashi et al, 2001; tochemical staining (Obama et al, 2005). The Hirohashi et al, 2002; Suzuki et al, 2002). frequency of expressed sequence tags of intrahepatic cholangiocarcinoma cell lines was GENE EXPRESSION PROFILE compared to normal liver tissues (Wang et AND CCA al, 2006). One hundred and thirty-seven genes are differentially expressed, and Cancer phenotype reflects changes in among them, ANXA1, ANXA2, AMBP and the expression patterns of hundreds or even SERPINC1 were verified in human ICC cell thousands of genes that occur as a conse- lines and tissues by semi-quantitative RT- quence of the primary mutation of an PCR and immunohistochemical analysis oncogene or a tumor suppressor gene. Much (Wang et al, 2006). of cancer research over the past 50 years has been devoted to analysis of genes that are METASTASIS- ASSOCIATED expressed differently in tumor cells as com- GENES IN CCA pared with their normal counter parts. To understand the molecular basis for To date, only a few studies of a compre- metastasis, a comparison of gene expression hensive analysis of gene expression in CCA profile of primary and metastatic tissues are have been reported. Global gene expression necessary. In several tumors, these metasta- analysis of primary resected tumors, biliary sis-associated genes have been defined cell lines and non-neoplastic biliary epithe- (Reyes et al, 2007; Basak et al, 2008; Stein et lial scrapings using Affymetrix oligonucle- al, 2009), but there are limited data of such otide microarrays has been conducted genes in CCA. (Hansel et al, 2003). Two hundred and eighty- two known genes are up-regulated three- Increased c-erbB-2 expression contri- fold or more in cancer compared to normal butes to the development of CCA into an ad- epithelium, including proliferation and cell vanced stage associated with tumor metasta- cycle antigens (cyclins D2 and E2, cdc2/p34, sis (Aishima et al, 2002). Expression of thy- and geminin), transcription factors midine phosphorylase, an important regu- (homeobox B7 and islet-1), growth factors lator of angiogenesis, shows a significant and growth factor receptors (hepatocyte correlation with vascular invasion, lym- growth factor, amphiregulin, and insulin- phatic permeation, perineural invasion and like growth factor 1 receptor), and enzymes lymph node metastasis of intrahepatic CCA modulating sensitivity to chemotherapeutic (Aishima et al, 2002). In addition, sialyl agents (cystathionine synthase, dCMP Lewis (a) is related to vascular invasion and deaminase, and CTP synthase). Gene expres- poor prognosis of CCA (Juntavee et al, 2005). sion profiles of intrahepatic CCA using tu- On the other hand, expression of bcl-2 is in- mor cell populations purified by laser mi- versely related to lymph node metastasis, crobeam microdissection were analyzed vascular invasion and perineural invasion (Obama et al, 2005). Fifty-two genes are com- of CCA (Ito et al, 2000). monly up-regulated and 421 down-regu- The incidence rate of lymph node me- lated in intrahepatic CCA compared with tastasis has been shown to increase in pro- noncancerous biliary epithelial cells. Among portion to an increase in the expression of the 52 up-regulated genes, P-cadherin and matrix metalloproteinase-9 in intrahepatic survivin were investigated for enhanced CCA (Shirabe et al, 1999). Over-expression

704 Vol 40 No. 4 July 2009 GENES AND CHOLANGIOCARCINOMA of aspartyl (asparaginyl)-hydroxylase is cor- therapeutic role, such as human MutL ho- related with cell motility of CCA (Maeda et molog1 (hMLH1), methionine aminopepti- al, 2003) and cyclin D1 over-expression is dase 2 (MetAP2), galectin-3, C-X-C motif more frequently observed in CCA with chemokine receptor 4 (CXCR4) and lymph node metastasis (Ito et al, 2001). Like- phosphoinositide-3-kinase (PI3K). wise, WISP1v expression is associated with Limpaiboon et al (2005) studied genetic lymphatic and perineural spread of CCA and epigenetic alterations of hMLH1 gene and poor clinical outcome (Tanaka et al, using polymerase chain reaction (PCR)-based 2003). microsatellite marker D3S1611 and methylation-specific PCR, respectively. hMLH1 is one CCA-RELATED GENES AND THEIR of the major DNA mismatch repair genes. POTENTIAL ROLES Hypermethylation of hMLH1 promoter oc- Several genes have been identified to be curs in 29 of 65 CCA cases (45%), while all associated with CCA and each gene has its normal tissues show unmethylation. Genetic own unique potential benefit. Table 1 sum- alterations of hMLH1 shown by the loss of marizes the study details of these genes heterozygosity (LOH) was detected in 12 of which are categorized as diagnostic, thera- 51 informative cases (23.5%) and microsatel- peutic and prognostic roles. lite instability (MSI) was found in 5 of 65 cases (8%). Of 29 cases showing hypermethylation Diagnostic role of CCA-related genes of hMLH1, 4 (14%) and 5 (17%) show MSI at Early diagnosis of CCA may provide D3S1611 and LOH, respectively. Even though better prognosis and survival. Srisomsap et hMLH1 promoter hypermethylation has no al (2007) compared the expression of mem- association with gross type, nerve invasion brane-associated and cytosolic proteins from and lymphatic invasion of CCA, it is signifi- human CCA (HuCCA-1) and hepatocellular cantly related to the poorly differentiated type carcinoma cell lines (HCC-S102) by 2-DE and (p = 0.013). These findings suggest that both LC/MS/MS. The proteomic analysis showed genetic and epigenetic mechanisms cause in- that the cytoskeletal microfilament-associ- activation of hMLH1. Since gene silencing by ated proteins were highly expressed in both methylation in an early event in carcinogen- membrane-associated and cytosolic proteins esis, promoter hypermethylation of hMLH1 of HuCCA-1 (40.9% in membrane and 26.2% may be a target of therapy and prevention of in cytosol) than in HCC-S102 (3.0% in mem- liver fluke- related CCA. brane and 9.1% in cytosol). In addition, ten Methionine aminopeptidases 2 (MetAP2), membrane proteins including calgizzarin, a metallopeptidase that selectively catalyzes ezrin, moesin, radixin, immunoglobulin the removal of the N-terminal methionine kappa chain variable region, integrin alpha- from newly synthesized protein, has been 6 precursor, cytochrome c oxidase polypep- reported to be related with tumor growth in tide VIb isoform 1, glycerol-3-phosphate several cancers (Datta and Datta, 1999; dehydrogenase, hippocalcin-like protein 1, Catalano et al, 2001; Selvakumar et al, 2006). and MAPK/ERK kinase 2 found only in There has been only one study of MetAP2 HuCCA-1, may be possible diagnostic mark- in CCA (Sawanyawisuth et al, 2007). Immu- ers for discrimination of CCA from HCC. nohistochemistry of MetAP2 was evaluated Therapeutic role of CCA-related genes in 59, 52 and 82 liver specimens with There are several potential genes that nonneoplasia, dysplasia and CCA, respec- have been shown to be associated with a tively. Fifty percent of the normal bile duct

Vol 40 No. 4 July 2009 705 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH is negative for MetAP2 and the remaining CCA cell invasion. These findings imply a expresses weak staining. However, hyper- potential role for the inhibition of CXCR4 or plastic/ dysplastic (62%) and primary tumor its signal cascades in the treatment of CCA. (73%) of biliary epithelial cells are uniformly In addition to the invasion effect, PI3K/ stained with moderate to strong intensity. Akt activation has been shown to be possible The expression level of MetAP2 is signifi- mechanism of chemotherapy-resistance of cantly increased in dysplastic and tumor bile CCA (Leelawat et al, 2009). The oxabilplatin- duct than in normal bile duct. Similar to the treated CCA cell lines (RMCCA1 and dysplastic and CCA bile duct epithelia, KKU100) show high levels of Akt and mam- MetAP2 is constantly expressed in meta- malian target of rapamycin (mTOR), a static tumors with moderate to high inten- downstream effector of PI3K/Akt activation. sity (Sawanyawisuth et al, 2007). However, The combination of oxaliplatin with there is no significant association of MetAP2 LY294001, an inhibitor of PI3K, results in a expression and pathological and clinical fea- remarkable arrest of cell proliferation and a tures (tumor grade, tumor stage, tumor size significant induction of apoptosis. Therefore, and tumor invasion). activation of PI3K might protect CCA cells Another treatment option to control from oxaliplatin regimen. Targeting the PI3K metastatic CCA has been shown by Junking pathway may be a useful treatment to im- et al (2008). Galectin-3 is significantly less ex- prove chemotherapeutic sensitivity of CCA. pressed in the poorly-differentiated CCA Prognostic role of CCA-related genes than in the papillary, well- to moderately- CCA-related genes that correlated with differentiated type (p =0.012). In addition, prognosis or survival have been reported low galectin-3 expression is significantly as- mainly from Khon Kaen University, Khon sociated with lymphatic invasion (p = 0.002). Kaen, Thailand, the endemic area of CCA Therefore, regulation of galectin-3 expres- (Limpaiboon et al, 2006; Muenphon et al, sion may be an alternative therapeutic ap- 2006; Thanasai et al, 2006; Thuwajit et al, 2008; proach to control metastasis of CCA. Chinnasri et al, 2009). Detection of specific Leelawat et al (2007, 2009), reported genes may indicate the aggressiveness of two potential mechanisms on treatment of CCA and also therapeutic interventions. CCA via C-X-C motif chemokine receptor Various study techniques and targeted mol- 4 (CXCR4) and its signaling cascades, which ecules or chromosomal regions are shown is a key factor for cancer cell progression in Table 1. and metastasis in CCA cell lines. The acti- Ten polymorphic microsatellite markers vation of CXCR4 with CXC chemokine were explored by LOH and MSI on chromo- ligand-12 (CXCL12) trigs the signaling via some region 1p36-pter in 90 CCA patients extracellular signal-regulated kinase-1/2 (Limpaiboon et al, 2006). Of those, 68 cases (ERK1/2) and phosphoinositide 3-kinase (75.6%) show LOH in one or more loci and (PI3K). To confirm this mechanism, the au- the most frequent loci are D1S199 (40.0%), thors found no invasion activity when CCA D1S507 (34.6%), D1S2845 (30.5%) and cell lines (RMCCA1 and KKU100) are pre- D1S2734 (30.1%). MSI was found in 34 cases treated with a specific inhibitor of CXCR4 (37.8%) at one or more loci. Patients with LOH (AMD3100) and then treated with CXCL12. at D1S234 and D1S2676 have significantly less In addition, treatment with MEK1/2 inhibi- and more lymphatic invasion, respectively tor (U0126) or PI3K inhibitor (LY294002) also (p = 0.017 and 0.031, respectively). LOH at attenuates the effect of CXCL12-induced D1S2845 shows a significant correlation with

706 Vol 40 No. 4 July 2009 GENES AND CHOLANGIOCARCINOMA

cinoma cell

MS and

2-DE, LC

PCR Immunohistochemistry

PCR, immunohistochemistry

oles.

egion, blotting Western

otein 1, and

ol-3-phosphate

Akt blotting Western

eceptor 4 blotting Western

INK4a

cinoma cell line-1; HCC-S102, hepatocellular car

ecursor, cytochrome c oxidase cytochrome ecursor,

, p16

INK4b

chromosomal regionchromosomal Technique

ERK kinase 2

, p15

eaction

ARF

able 1

calgizzarin, ezrin, moesin, radixin, immunoglobulin kappa chain variable r integrin alpha-6 pr polypeptide VIb isoform 1, glycer dehydrogenase, hippocalcin-like pr dehydrogenase, MAPK

Human mutL homolog1 Methionine aminopeptidase 2 C-X-C motif chemokine r Galectin-3 Immunohistochemistry

Phosphoinositide-3-kinase

1p36-pter PCR

p14

KKU 100, KKU214

Cholangiocarcinoma-related genes categorized by their potential r Cholangiocarcinoma-related

cinoma patients; PCR, polymerase chain r

Srisomsap 2007 NA HUCCA-1, HCC-S102

Limpaiboon 2005 65 Intrahepatic CCA Sawanyawisuth 2007 82 Intrahepatic CCA Leelawat 2007 NA RMCCA1, KKU100 Junking 2008 53 Intrahepatic CCA,

Leelawat 2009 NA RMCCA1, KKU100

Limpaiboon 2006 90 Intrahepatic CCA Muenphon 2006 80 CCA 21q22-qter PCR Quantitative real-time Thanasai 2006 65 CCA 22q12-qter PCR Quantitative real-time Thuwajit 2008 51 CCA peptide trefoil Mucin6, trefoil-2 Immunohistochemistry Chinnasri 2009 94 Intrahepatic CCA

Prognostic

Study Year N Source Molecule

Diagnostic

Therapeutic

N, numbers of subjects; NA, not applicable; HUCCA-1, human bile duct epithelial car

line S102; CCA, cholangiocar

Vol 40 No. 4 July 2009 707 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH nerve invasion and MSI at D1S228 has a poor patients by immunohistochemistry. The high prognosis (p = 0.029 and 0.0026, respectively). expression levels of both proteins signifi- The authors concluded that allelic loss at chro- cantly correlate with prolonged post-opera- mosome 1p36 can be used as molecular prog- tive survival time, but only a high expres- nostic indicators for CCA patients. sion of MUC6 is significantly related with a Researchers from the Liver Fluke and 5-year survival rate. Independent poor prog- Cholangiocarcinoma Research Center, Khon nostic factors including a low expression of Kaen University, Khon Kaen, Thailand also MUC6, high expression of TFF2, age of pa- discovered other prognostic markers for tients of more than 56 years, tumor size of CCA. D21S1893-D21S1890 region of chromo- more than 5 cm, and poorly-differentiated some 21q22-qter may carry candidate genes histological type were indicated by multi- especially trefoil factor or serine protease variate Cox regression analysis. family and might be involved in tumor invasion, metastasis and poor survival CONCLUSION (Muenphon et al, 2006). On the other hand, CCA is a slow growing cancer with poor D22S283 amplification on 22q12-qter chro- prognosis. It is highly prevalent in North- mosomal area is an independent indicator east Thailand where it is endemic. In this of favorable prognosis in liver fluke- related region, CCA is associated with OV infection. CCA (Thanasai et al, 2006). Several researchers aimed to improve the Chinnasri et al (2009) showed a correla- procedures of diagnosis and treatment by tion between clinicopathological data and targeting molecule markers related to CCA. genetic-epigenetic alteration including loss This review indicates the significance of ARF INK4b of protein expressions of p14 , p15 , and CCA-related genes and their potential diag- INK4a p16 in 94 CCA samples. LOH, MSI and nostic, therapeutic and prognostic roles. fine mapping of the chromosomal region Clinical application of these molecular mark- 9p21-pter were performed using PCR-based ers may be useful for early detection and microsatellite markers. Methylation and pro- therapy of CCA in the future. However, fur- ARF INK4b tein expression of p14 , p15 , and ther studies are still needed to provide pre- INK4a p16 were determined using methylation- vention strategies and the best therapeutic specific PCR and immunohistochemistry, outcome. respectively. Region between D9S286 and D9S1752 of chromosome 9p21-pter was demonstrated to be a common loss. Methy- ACKNOWLEDGEMENTS ARF INK4b lation frequency for p14 , p15 , and The author thanks Dr Kittisak Sawanya- INK4a p16 is 40.2, 48.9, 28.3%, respectively. Loss wisuth for his kind support and assistance. of these 3 proteins is 30.9, 58, and 81.5%, respectively. Poor prognosis is associated with only the loss of p16INK4a by Kaplan- REFERENCES Meier survival analysis (p = 0.026). Aishima SI, Taguchi KI, Sugimachi K, Shimada Additionally, mucin6 (MUC6), a mucin M, Tsuneyoshi M. c-erbB-2 and c-Met ex- that is normally co-expressed with the tre- pression relates to cholangiocarcinogenesis foil factor family-2 (TFF2) trefoil peptide, has and progression of intrahepatic cholangio- shown a good correlation with survival of carcinoma. Histopathology 2002; 40: 269-78. CCA patients (Thuwajit et al, 2008). MUC6 Anderson CD, Pinson CW, Berlin J, Chari RS. Di- and TFF2 levels were examined in 51 CCA agnosis and treatment of cholangiocarci-

708 Vol 40 No. 4 July 2009 GENES AND CHOLANGIOCARCINOMA

noma. Oncologist 2004; 9: 43-57. clinicopathologic features and surgical out- Bartsch H, Ohshima H, Pignatelli B, Calmels S. comes. Hepatogastroenterology 2002; 49: 326- Endogenously formed N-nitroso com- 9. pounds and nitrosating agents in human Hou PC. The relationship between primary car- cancer etiology. Pharmacogenetics 1992; 2: cinoma of the liver and infestation with 272-7. clonorchiasis sinensis. J Path Bact 1956; 72: Basak S, Jacobs SB, Krieg AJ, et al. The metasta- 239-46. sis-associated gene Prl-3 is a p53 target in- Ito Y, Takeda T, Sasaki Y, et al. Bcl-2 expression in volved in cell-cycle regulation. Mol Cell 2008; cholangiocellular carcinoma is inversely 30: 303-14. correlated with biologically aggressive phe- Carriaga MT, Henson DE. Liver, gallbladder, ex- notypes. Oncology 2000; 59: 63-7. trahepatic bile ducts, and pancreas. Cancer Ito Y, Takeda T, Sasaki Y, et al. Expression and 1995; 75: 171-90. clinical significance of the G1-S modulators Catalano A, Romano M, Robuffo I, Strizzi L, in intrahepatic cholangiocellular carcinoma. Procopio A. Methionine aminopeptidase-2 Oncology 2001; 60: 242-51. regulates human mesothelioma cell sur- Junking M, Wongkham C, Sripa B, Sawanya- vival: role of Bcl-2 expression and wisuth K, Araki N, Wongkham S. Decreased telomerase activity. Am J Pathol 2001; 159: expression of galectin-3 is associated with 721-31. metastatic potential of liver fluke- associated Chinnasri P, Pairojkul C, Jearanaikoon P, et al. cholangiocarcinoma. Eur J Cancer 2008; 44: Preferentially different mechanisms of inac- 619-26. tivation of 9p21 gene cluster in liver fluke- Juntavee A, Sripa B, Pugkhem A, Khuntikeo N, related cholangiocarcinoma. Hum Pathol Wongkham S. Expression of sialyl Lewis(a) 2009 [Epub ahead of print]. relates to poor prognosis in cholangiocarci- Datta B, Datta R. Induction of apoptosis due to noma. World J Gastroenterol 2005; 11: 249-54. lowering the level of eukaryotic initiation Khan SA, Taylor-Robinson SD, Toledano MB, factor 2-associated protein, p67, from mam- Beck A, Elliott P, Thomas HC. Changing in- malian cells by antisense approach. Exp Cell ternational trends in mortality rates for liver, Res 1999; 246: 376-83. biliary and pancreatic tumours. J Hepatol Farley DR, Weaver AL, Nagorney DM. “Natural 2002; 37: 806-13. history” of unresected cholangiocarcinoma: Khan SA, Toledano MB, Taylor-Robinson SD. patient outcome after noncurative interven- Epidemiology, risk factors, and pathogen- tion. Mayo Clin Proc 1995; 70: 425-9. esis of cholangiocarcinoma. HPB (Oxford) Gores GJ. Cholangiocarcinoma: current concepts 2008; 10: 77-82. and insights. Hepatology 2003; 37: 961-9. Khuntikeo N, Pugkhem A, Bhudhisawasdi V, Hanazaki K, Kajikawa S, Shimozawa N, et al. Uttaravichien T. Major hepatic resection for Prognostic factors of intrahepatic cholan- hilar cholangiocarcinoma without preopera- giocarcinoma after hepatic resection: tive biliary drainage. Asian Pac J Cancer Prev univariate and multivariate analysis. 2008; 9: 83-5. Hepatogastroenterology 2002; 49: 311-6. Lee YM, Kaplan MM. Primary sclerosing cholan- Hansel DE, Rahman A, Hidalgo M, et al. Identifi- gitis. N Engl J Med 1995; 332: 924-33. cation of novel cellular targets in biliary tract Leelawat K, Leelawat S, Narong S, Hongeng S. cancers using global gene expression tech- Roles of the MEK1/2 and AKT pathways in nology. Am J Pathol 2003; 163: 217-29. CXCL 12/CXCR4 induced cholangiocarci- Hirohashi K, Uenishi T, Kubo S, et al. Macroscopic noma cell invasion. World J Gastroenterol types of intrahepatic cholangiocarcinoma: 2007; 13: 1561-8.

Vol 40 No. 4 July 2009 709 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

Leelawat K, Narong S, Udomchaiprasertkul W, sis of gene expression in human intrahepatic Leelawat S, Tungpradubkul S. Inhibition of cholangiocarcinoma. Hepatology 2005; 41: PI3K increases oxaliplatin sensitivity in 1339-48. cholangiocarcinoma cells. Cancer Cell Int Ohshima H, Bandaletova TY, Brouet I, et al. In- 2009; 9: 3. creased nitrosamine and nitrate biosynthe- Limpaiboon T, Khaenam P, Chinnasri P, et al. Pro- sis mediated by nitric oxide synthase in- moter hypermethylation is a major event of duced in hamsters infected with liver fluke MLH1 gene inactivation in liver fluke re- (Opisthorchis viverrini). Carcinogenesis 1994; lated cholangiocarcinoma. Cancer Lett 2005; 15: 271-5. 217: 213-9. Okabayashi T, Yamamoto J, Kosuge T, et al. A new Limpaiboon T, Tapdara S, Jearanaikoon P, Sripa staging system for mass-forming intrahe- B, Bhudhisawasdi V. Prognostic significance patic cholangiocarcinoma: analysis of pre- of microsatellite alterations at 1p36 in operative and postoperative variables. Can- cholangiocarcinoma. World J Gastroenterol cer 2001; 92: 2374-83. 2006; 12: 4377-82. Pairojkul C, Shirai T, Hirohashi S, et al. Multistage Liu D, Momoi H, Li L, Ishikawa Y, Fukumoto M. carcinogenesis of liver-fluke-associated Microsatellite instability in thorotrast-in- cholangiocarcinoma in Thailand. Princess duced human intrahepatic cholangiocarci- Takamatsu Symp 1991; 22: 77-86. noma. Int J Cancer 2002; 102: 366-71. Parkin DM, Ohshima H, Srivatanakul P, Maeda T, Sepe P, Lahousse S, et al. Antisense Vatanasapt V. Cholangiocarcinoma: epide- oligodeoxynucleotides directed against as- miology, mechanisms of carcinogenesis and partyl (asparaginyl) beta-hydroxylase sup- prevention. Cancer Epidemiol Biomarkers Prev press migration of cholangiocarcinoma cells. 1993; 2: 537-44. J Hepatol 2003; 38: 615-22. Patel T. Increasing incidence and mortality of Mitacek EJ, Brunnemann KD, Hoffmann D, et al. primary intrahepatic cholangiocarcinoma in Volatile nitrosamines and tobacco-specific the United States. Hepatology 2001; 33: 1353- nitrosamines in the smoke of Thai cigarettes: 7. a risk factor for lung cancer and a suspected Pinlaor S, Ma N, Hiraku Y, et al. Repeated infec- risk factor for liver cancer in Thailand. Car- tion with Opisthorchis viverrini induces ac- cinogenesis 1999; 20: 133-7. cumulation of 8-nitroguanine and 8-oxo-7,8- Muenphon K, Limpaiboon T, Jearanaikoon P, dihydro-2'-deoxyguanine in the bile duct of Pairojkul C, Sripa B, Bhudhisawasdi V. hamsters via inducible nitric oxide synthase. Amplification of chromosome 21q22.3 har- Carcinogenesis 2004; 25: 1535-42. boring trefoil factor family genes in liver Pinlaor S, Yongvanit P, Hiraku Y, et al. 8- fluke related cholangiocarcinoma is associ- nitroguanine formation in the liver of ham- ated with poor prognosis. World J sters infected with Opisthorchis viverrini. Gastroenterol 2006; 12: 4143-8. Biochem Biophys Res Commun 2003; 309: 567- Nakeeb A, Pitt HA, Sohn TA, et al. Cholangio- 71. carcinoma. A spectrum of intrahepatic, Rosen CB, Nagorney DM, Wiesner RH, Coffey perihilar, and distal tumors. Ann Surg 1996; RJ Jr, LaRusso NF. Cholangiocarcinoma 224: 463-73. complicating primary sclerosing cholangi- Namieno T, Koito K, Takahashi M, Une Y, tis. Ann Surg 1991; 213: 21-5. Yamashita K, Shimamura T. Survival-asso- Reyes I, Tiwari R, Geliebter J, Reyes N. DNA ciated histologic spreading modes of oper- microarray analysis reveals metastasis-asso- able intrahepatic, peripheral-type cholan- ciated genes in rat prostate cancer cell lines. giocarcinomas. World J Surg 2001; 25: 572-7. Biomedica 2007; 27: 190-203. Obama K, Ura K, Li M, et al. Genome-wide analy- Satarug S, Haswell-Elkins MR, Tsuda M, et al.

710 Vol 40 No. 4 July 2009 GENES AND CHOLANGIOCARCINOMA

Thiocyanate-independent nitrosation in identified key regulator of HGF-MET sig- humans with carcinogenic parasite infec- naling, predicts colon cancer metastasis. Nat tion. Carcinogenesis 1996; 17: 1075-81. Med 2009; 15: 59-67. Sawanyawisuth K, Wongkham C, Pairojkul C, et Suzuki S, Sakaguchi T, Yokoi Y, et al. Clinico- al. Methionine aminopeptidase 2 over-ex- pathological prognostic factors and impact pressed in cholangiocarcinoma: potential for of surgical treatment of mass-forming intra- drug target. Acta Oncol 2007; 46: 378-85. hepatic cholangiocarcinoma. World J Surg Selvakumar P, Lakshmikuttyamma A, Dimmock 2002; 26: 687-93. JR, Sharma RK. Methionine aminopeptidase Tanaka S, Sugimachi K, Kameyama T, et al. Hu- 2 and cancer. Biochim Biophys Acta 2006; man WISP1v, a member of the CCN family, 1765: 148-54. is associated with invasive cholangiocarcinoma. Hepatology 2003; 37: 1122-9. Shirabe K, Shimada M, Harimoto N, et al. Intra- hepatic cholangiocarcinoma: its mode of Taoka H, Kawarada Y. Intrahepatic bile duct car- spreading and therapeutic modalities. Sur- cinoma (cholangiocarcinoma). Nippon Geka gery 2002; 131: S159-64. Gakkai Zasshi 1997; 98: 484-90. Shirabe K, Shimada M, Kajiyama K, et al. Expres- Thamavit W, Pairojkul C, Tiwawech D, Itoh M, sion of matrix metalloproteinase-9 in surgi- Shirai T, Ito N. Promotion of cholangiocarci- cally resected intrahepatic cholangiocarcinogenesis in the hamster liver by bile duct noma. Surgery 1999; 126: 842-6. ligation after dimethylnitrosamine initiation. Carcinogenesis 1993; 14: 2415-7. Sriamporn S, Pisani P, Pipitgool V, Suwan- rungruang K, Kamsa-ard S, Parkin DM. Thamavit W, Pairojkul C, Tiwawech D, Shirai T, Prevalence of Opisthorchis viverrini infection Ito N. Strong promoting effect of and incidence of cholangiocarcinoma in Opisthorchis viverrini infection on dimeth- Khon Kaen, Northeast Thailand. Trop Med ylnitrosamine-initiated hamster liver. Can- Int Health 2004; 9: 588-94. cer Lett 1994; 78: 121-5. Sripa B, Kaewkes S, Sithithaworn P, et al. Liver Thanasai J, Limpaiboon T, Jearanaikoon P, et al. fluke induces cholangiocarcinoma. PLos Amplification of D22S283 as a favorable Med 2007; 4: e201. prognostic indicator in liver fluke related cholangiocarcinoma. World J Gastroenterol Sripa B, Pairojkul C. Pathology of cholangiocar- 2006; 12: 4338-44. cinoma. In: Vatanasapt V, Sripa B, eds. Liver cancer in Thailand; epidemiology, diagno- Thuwajit C, Thuwajit P, Kaewkes S, et al. In- sis and control. Khon Kaen: Siriphan Press creased cell proliferation of mouse fibroblast / 2000: 65-100. NIH-3T3 in vitro induced by excretory secretory product(s) from Opisthorchis viverrini. Sripa B, Pairojkul C. Cholangiocarcinoma: les- Parasitology 2004; 129: 455-64. sons from Thailand. Curr Opin Gastroenterol 2008; 24: 349-56. Thuwajit C, Thuwajit P, Uchida K, et al. Gene expression profiling defined pathways cor- Srisomsap C, Subhasitanont P, Sawangareetrakul related with fibroblast cell proliferation in- P, et al. Comparison of membrane-associated duced by Opisthorchis viverrini excretory/ proteins in human cholangiocarcinoma and secretory product. World J Gastroenterol 2006; hepatocellular carcinoma cell lines. 12: 3585-92. Proteomics Clin Appl 2007; 1: 89-106. Thuwajit P, Chawengrattanachot W, Thuwajit C, Srivatanakul P, Ohshima H, Khlat M, et al. En- Sripa B, Paupairoj A, Chau-In S. Enhanced dogenous nitrosamines and liver fluke as expression of mucin 6 glycoprotein in risk factors for cholangiocarcinoma in Thai- cholangiocarcinoma tissue from patients in land. IARC Sci Publ 1991; 105: 88-95. Thailand as a prognostic marker for sur- Stein U, Walther W, Arlt F, et al. MACC1, a newly vival. J Gastroenterol Hepatol 2008; 23: 771-8.

Vol 40 No. 4 July 2009 711 SOUTHEAST ASIAN J TROP MED PUBLIC HEALTH

Uenishi T, Hirohashi K, Kubo S, Yamamoto T, noma in Thailand. J Hepatobiliary Pancreat Yamazaki O, Kinoshita H. Clinicopathologi- Surg 1999; 6: 128-35. cal factors predicting outcome after resection Wang AG, Yoon SY, Oh JH, et al. Identification of of mass-forming intrahepatic cholangio- intrahepatic cholangiocarcinoma related carcinoma. Br J Surg 2001; 88: 969-74. genes by comparison with normal liver tis- Uttaravichien T, Bhudhisawasdi V, Pairojkul C, sues using expressed sequence tags. Biochem Pugkhem A. Intrahepatic cholangiocarci- Biophys Res Commun 2006; 345: 1022-32.

712 Vol 40 No. 4 July 2009