Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Aug 9.
Published in final edited form as: Int J Parasitol. 2010 May 21;40(10):1203–1212. doi: 10.1016/j.ijpara.2010.05.002

Up-regulation of annexin A2 in cholangiocarcinoma caused by Opisthorchis viverrini and its implication as a prognostic marker

Ponlapat Yonglitthipagon a, Chawalit Pairojkul a,b, Yaovalux Chamgramol a,b, Jason Mulvenna d, Banchob Sripa a,b,c,*
PMCID: PMC3739719  NIHMSID: NIHMS215024  PMID: 20493868

Abstract

Cholangiocarcinoma (CCA), or cancer of the bile ducts, is primarily associated with infection with the liver fluke Opisthorchis viverrini in northeast Thailand. The disease . is associated with late presentation, poses challenges for diagnosis and has a high mortality rate – features that highlight the need for tumor markers. At present, there are no specific tumor markers that can indicate the early stages and status of CCA. Proteomic analysis of the proteins expressed on the surface of tumor cells is particularly difficult since proteome-wide analysis of surface membrane proteins has thus far been hampered by the lack of effective strategies to profile hydrophobic membrane proteins. In this study, a sequential protein extraction was utilized to overcome this problem. Membrane protein was extracted from four CCA cell lines with different tumor forming capabilities. The non-tumor H69 biliary cell line was used as a control. Two-dimensional-PAGE followed by MALDI-TOF-MS was used to identify differentially expressed proteins. Among 20 up-regulated membrane proteins identified in the CCA cell lines was ANXA2, a participant in tumor invasion and metastasis in other cancers. Accordingly, ANXA2 was verified in human subjects by probing, using a commercial anti-mouse monoclonal antibody and a tissue microarray of CCA (301 diagnosed cases), where it was found to associate with one of several tumor progression stages as reflected by lymphatic invasion (P = 0.014) and metastasis (P = 0.026). Patients with high expression of ANXA2 had a significantly shorter survival time (P = 0.011). ANXA2 expression in tumors may be useful for predicting the poor outcome of CCA patients.

Keywords: Opisthorchis viverrini, Annexin, ANXA2, Cholangiocarcinoma, Proteomics, Prognotic marker

1. Introduction

Cholangiocarcinoma (CCA) is a primary cancer of the bile ducts. In Thailand, the highest incidence of CCA is found in the northeastern region and is primarily associated . with infection with the liver fluke Opisthorchis viverrini,a food-borne trematode which chronically infects the bile ducts of human beings. People become infected by eating raw or undercooked freshwater fish in dishes such as koi pla which contain the metacercaria. Immature flukes migrate up through the ampulla of Vater to the biliary tree and mature in the small intrahepatic ducts (Sripa and Pairojkul, 2008). An estimated 6 million people are infected with O. viverrini in Thailand. The geographical pattern of liver fluke infection is not uniform, however, with the greatest prevalence in the north (19.3%) and northeast (15.7%) compared with the central (3.8%) and southern regions (0%). Despite wide-spread treatment with praziquantel, the prevalence of O. viverrini in some endemic areas approaches 70% (Jongsuksuntigul and Imsomboon, 2003). It has long been known that infection with O. viverrini is associated with the induction of CCA (Sriamporn et al., 2004; Sriamporn et al., 2005; Sripa et al., 2007), indeed there is no stronger link between an eukaryotic organism and a malignant neoplasm than that between O. viverrini and CCA. This has led the World Health Orgainization’s International Agency for Research on Cancer to classify O. viverrini as a Group 1 carcinogen to humans (Bouvard et al., 2009; IARC, 2009). For unknown reasons, cases of CCA are increasing worldwide (Blechacz and Gores, 2008). Only the surgical resection of all detectable tumors leads to an improvement in the 5-year survival rate. However, a complete resection is often impossible, typically resulting in subsequent metastasis and local recurrence (Olnes and Erlich, 2004). At present, there are no specific tumor markers that can indicate the early stages and status of CCA. Thus, there is a need for tumor markers than can be measured early and in accessible samples such as plasma.

Mass spectrometry (MS) is a powerful approach for the identification of differentially expressed, low-abundance proteins that may be useful as tumor markers. Proteomic approaches have been successfully employed in studies of breast (Kischel et al., 2008) and prostate (Sardana et al., 2007) cancer for tumor marker discovery. Of particular interest in these studies are the proteins expressed on the surface of tumor cells and characterization of cell surface proteomes could provide a better understanding of the manner in which those are regulated. However, proteome-wide analysis of surface membrane proteins has thus far been hampered by the lack of effective strategies to profile hydrophobic membrane proteins. A method to overcome this problem, sequential protein extraction combined with Tris, DTT and Tributylphosphine (TBP), has recently been utilized to isolate and identify surface proteins from human breast cancer cell lines (Ruan and Wan, 2007). In the present work, we interrogated the cell membranes of human CCA cell lines using two-dimensional gel electrophoresis (2-DE) in order to identify proteins associated with the plasma membrane, and to understand the cellular abundance and status of these proteins. This has allowed us to detect tumor markers for CCA, with one protein of particular interest, annexin A2 (ANXA2), identified by MALDI-MS. Over-expression of ANXA2 has been directly implicated in liver cancer progression (Mohammad et al., 2008; Ji t al., 2009), and here we show its utility as a potential prognostic marker for O. viverrini-induced CCA.

2. Materials and methods

2.1. Cell cultures

Four human CCA cell lines, namely KKU-M156 (M156), KKU-100 (K100), KKU-139 (M139) and KKU-M213 (M213), were isolated from CCA patients from northeastern Thailand. Cell lines were developed and cultured as described elsewhere (Sripa et al., 2005). In all cases, CCA was linked to infection with O. viverrini. Approval for use of the tissue was obtained from the Human Research Ethics Committee of Khon Kaen University, Thailand. CCA tissues were histologically characterized based on their . marginal zones as moderately differentiated adenocarcinoma (M156), poorly differentiated adenocarcinoma (K100), squamous cell carcinoma (M139) and adenosquamous cell carcinoma (M213). CCA cell lines were cultured in RPMI-1640 medium (Gibco, Grand Island, NY, USA) containing 100 U/mL penicillin and 100 µg/mL streptomycin with 10% FBS (Hyclone Laboratories). H69 cells, a human ‘immortalized’, non-malignant cholangiocyte cell line, were cultured as previously described (Grubman et al., 1994). Cell growth was performed at 37°C under 5% CO2 and 95% humidified air. During this experiment mycoplasma contamination was periodically checked and no contamination was present in any cell line.

2.2. Clinical tissue samples

In this study, CCA tissues were obtained with informed consent from the patients who underwent hepatectomy at Srinagarind hospital, Khon Kaen University, Thailand. The experimental protocol was approved by the Human Research Ethics committee of Khon Kaen University (IRB number HE42075). All tumors were clinically and histologically diagnosed as intrahepatic CCA (ICC) according to WHO classifications (Nakanuma et al., 2000). Patient information was obtained from medical records. Infection with O. viverrini in the patients was diagnosed as positive if the patient fulfilled one of these four criteria: (i) a history of previous positive stool examination for O. viverrini or its eggs; (ii) identification of O. viverrini or its eggs in stool or bile; (iii) demonstration of a typical bead-like cholangiogram; or (iv) histological evidence of O. viverrini in the specimen (Watanapa, 1996). Of the 301 liver fluke-associated CCA patients explored, 203 were male and 98 were female with a male to female ratio of 2:1. The mean age was 55 ± 9 years (range, 31–75 years). Most of the patients were at an advanced CCA stage, with 73.9% lymphatic, 53.1% vascular and 39.6% neural invasion. Thirty-five percent of the . tumors had well-differentiated histopathological grading. The majority of patients (63.5%) possessed a tumor size > 5 cm.

2.3. Sample preparation

Cells were examined under a phase-contrast microscope to ensure that they were 90–95% confluent before lysis. The culture medium was discarded and the cells were washed with 0.25 M sucrose three times on ice. Cells were scraped thoroughly with a scraper in 0.25 M sucrose containing 1% Protease Inhibitor Mix (Amersham Biosciences). The cells were collected and centrifuged at 1,500 g for 5 min at 4°C. The pellets were resuspended in lysis buffer containing 7 M urea, 2 M thiourea, 4% CHAPS (zwitterionic detergent commonly used in isoelectric focusing), 2% immobilized pH gradient (IPG) buffer 3–10 NL, 40 mM DTT, and 1% Protease Inhibitor Mix and allowed to lyse on ice for 15 min. Cell lysates were sonicated according to the manufacturer’s instructions (Sonics & Materials Inc. VCX 400, USA) and incubated for 2 h at 4°C. The lysate was then centrifuged at 600 g for 10 min to remove the nuclei and unlysed cells. The supernatant was centrifuged at 17,000 g for 30 min to harvest crude membrane fractions, then washed twice in ice-cold 0.25 M sucrose and centrifuged at 17,000 g for another 30 min. The membrane-enriched fractions were then resuspended in extraction buffer (7 M urea, 2 M thiourea, 4% CHAPS, 1% IPG buffer 3–10 NL, 50 mM DTT, 20 mM Tris, 0.5% Triton X-100, and 1% Protease inhibitor mix) at 4°C and were subjected to vortexing every 5 min for 1 h followed by centrifugation at 20,000 g for 1 h. The supernatant was collected and protein concentration was determined by Bradford Assay using a VersaMax™ absorbance microplate reader (Molecular Devices Corporation California, USA) at 595 nm. Membrane proteins of each cell line were then ready for 2-DE.

2.4. 2-DE and image analysis

2-DE was performed using the Immobiline/polyacrylamide system. For active rehydration loading, the strips were rehydrated with 125 µL rehydration buffer (7 M urea, 2 M thiourea, 4% CHAPS, 0.5% Triton X-100, 0.5% IPG buffer, 5 mM TBP with 30 mM DTT) containing 100 µg protein and applied to 7 cm pH 3–10 NL IPG strips (Amersham Biosciences, Sweden) by in-gel rehydration (20 V, 12 h), after which isoelectric focusing was performed for 17,840 Vh. IPG strips were then incubated in equilibration buffer (50 mM Tris, pH 8.8, 6 M urea, 30% glycerol, 10% SDS with a trace of bromophenol blue) containing 1% DTT for 15 min followed by 15 min incubation in equilibration buffer supplemented with 2.5% iodoacetamide. The IPG strips were then applied to the second-dimensional 12.5% SDS-polyacrylamide gels. Electrophoresis of the mini-gel was performed in a Hoefer system at 20 mA at room temperature for 2 h. The proteins were fixed in the gel and visualized by CBR-250 (GE Healthcare) staining. Stained gels were scanned using an ImageScanner (Amersham Biosciences, Sweden) and analyzed using ImageMaster™ 2D Platinum 6.0 software (GE Healthcare).

2.5. Tryptic in-gel digestion of 2-DE spots, MS analysis and protein identification

Protein spots that were unique to or stained more intensely in CCA cell lines compared with H69 cells were excised from the gel and transferred to V-bottom 96-well microtitre plates. Tryptic digestions were performed on an EttanTM Spot Handling Workstation robot (Amersham Biosciences) according to the manufacturer’s specifications. The 10 mg of matrix solution, α-cyano-4-hydroxycinnamic acid (Bruker Daltonik GmbH, Germany), was prepared by dissolving to saturation in 50% acetronitrile/water with 0.1% trifluoroacetic acid and then mixed with equal volumes of tryptic peptide with matrix solution. The mixture (1 µl) was spotted onto a steel target surface (MTP 384 ground steel, Bruker Daltonik) and dried by air. Mass spectra were recorded on an Autoflex MALDI-TOF mass spectrometer (Bruker Daltonik) with delayed extraction, 2GHz LeCroy digitizer and a 337 nm N2 laser. The MALDI-TOF MS was calibrated using trypsin auto-digestion peptide signals and matrix ion signals. Typically, 100 shots were accumulated from three to five different positions within a sample spot. The mass spectra were analyzed using MALDI evaluation software (Amersham Biosciences). Proteins were identified by peptide mass fingerprinting using Mascot (http://www.matrixscience.com) in searches against the NCBI non-redundant protein database (NCBInr). Parameters used in Mascot searches were as follows: taxonomy was restricted to Homo sapiens, trypsin was specified with allowance for one missed cleavage, a 2 Da peptide tolerance was allowed, and carbamidomethyl and oxidized methionine were chosen as the fixed and variable modifications, respectively. A threshold of 5% probability (P-value < 0.05) of a false positive was applied for all Mascot searches.

2.6. Western blot analyses

After 2-DE proteins were transferred to nitrocellulose membrane. The membrane was blocked with 5% skimmed milk in PBS/0.05% Tween-20 (PBS-T) for 2 h. The membrane was incubated overnight with mouse anti-Annexin II (SKU 03–4400, Invitrogen) diluted 1:1,000 (v/v) in 2% skimmed milk in PBS-T at 4°C. After washing 3 × 5 min with PBS-T the membrane was incubated in horseradish peroxidase (HRP)-conjugated goat anti-mouse antibody (Zymed Laboratories) diluted 1:2,000 (v/v) in 2% skimmed milk in PBS-T at room temperature for 1 h. The membrane was washed again for 3 × 5 min with PBS-T followed by a final 10 min rinse with PBS. Reactive spots were visualized with diaminobenzidine (DAB, Sigma Chemical Co).

2.7. Tissue microarray (TMA) and immunohistochemistry (IHC)

Tissue microarrays (TMAs) were constructed by the Department of Pathology, Faculty of Medicine, Khon Kaen University, Thailand. Following TMA construction (Fedor and De Marzo, 2005), an H&E stained section of the TMA recipient block was prepared and reviewed to confirm the presence of intact neoplasm. The arrays contained a total of 301 O. viverrini-associated CCA cases. One core was taken from each formalin-fixed, paraffin-embedded CCA sample by using punch cores that measured 0.9 mm in greatest diameter from the non-necrotic area of tumor foci.

Immunohistochemical reactions were performed on 4 µm-thick sections of TMA silane-coated slides (Sigma, St. Louis, MO, USA) by an immunoperoxidase method. TMA sections were deparaffinized in xylene and rehydrated in serial graded ethanol and distilled water, respectively. Antigen retrieval was performed by immersion of tissue sections in pre-heated 10 mM, pH 6.0 citrate buffer and maintaining heat at 120°C in a pressure cooker for 5 min. Endogenous peroxidase activity was eliminated by treating sectioned tissues with absolute methanol containing 3% H2O2 for 30 min. The sections were then washed in water and PBS. Non-specific staining was blocked by treating slides with 5% normal horse serum in PBS for 30 min. The sections were probed with mouse anti-Annexin II (as described earlier) diluted 1:400 (v/v) in PBS and incubated overnight at 4°C. After rinsing for 3 × 5 min with PBS the sections were incubated with HRP-conjugated goat anti-mouse IgG (Abcam Inc., USA) for 1 h. Sections were rinsed with PBS for 2 × 10 min, after which the sections were developed with DAB (Sigma Chemical Co.). The sections were counterstained with Mayer’s haematoxylin, dehydrated, cleared in xylene and mounted in Permount.

2.8. Assessment of IHC staining and statistical analyses

Immunoreactivity was evaluated independently by three researchers (Ponlapat Yonglitthipagon, Banchob Sripa and Chawalit Pairojkul) who were blinded to patient outcome. Consensus was reached whenever disagreement occurred. The percentage of positive tumor cells was determined using interactive stereological immunoscoring based on systematic random sampling (van Diest et al., 1997) and the average score was calculated. As show in Fig. 1, within a defined area of the lesion, we randomly selected five high-power fields (magnification, x400; 100 cells/high-power field) and approximately 500 tumor cells were counted. In this study the percentage of positive cells expressing ANXA2 was categorized: <10% (−), ≥ 10% (+) as described elsewhere (Zhuang et al., 2008). The extent of invasion by the cancer was determined in both tumorous and non-tumorous liver tissues by a senior pathologist (Chawalit Pairojkul). Lymphatic and vascular invasion was detected by the presence of infiltrating cancer cells within the lymphatic or blood vessels, respectively. Neural invasion was presented as positive cancer cells in the perineurium and/or neural fascicle.

Fig. 1.

Fig. 1

Open in a new tab

Schematic explanation of the protocol for interactive stereological immunoscoring based on systematic random sampling. Within a defined area of each tumor core in tissue microarray (TMA) of human Opisthorchis viverrini-associated CCA tissues, we randomly selected five high-power fields (magnification, ×400; 100 cells/high-power field) and approximately 500 tumor cells were counted.

Statistical analyses of the data was performed with the SPSS version 16.0 statistical package. For cross-sectional analyses, the chi-square test was utilized to analyze the relation between ANXA2 expression and categorical variables regarding clinical pathology parameters. Patient survival was calculated from the time of surgical resection to either the date of death regardless of cause or the most recent contact. The survivors were censored at the date they were last known to be alive. The Kaplan–Meier method was used to calculate cumulative survival. Deaths from causes other than the index tumor or its metastases were not considered to be treatment failures and were excluded from the final analyses of specific survival for the illness. Differences in survival between ANXA2 (−) and AXNA2 (+) groups were analyzed for significance by the log-rank method. The Cox-regression model was used to perform multivariate analysis and values of P < 0.05 were considered statistically significant. A schematic representation of the methods outlined is summarized in Fig. 2.

Fig. 2.

Fig. 2

Open in a new tab

Overview of combined proteomics and tissue microarray (TMA) procedures. Membrane proteins were extracted from cholangiocarcinoma (CCA) cell lines and separated by two-dimensional gel electrophoresis (2-DE). After image analysis, gel spots were excised and subjected to in-gel tryptic digestion. Peptides were analyzed by MALDI-TOF-MS. Peptide mass fingerprint (PMF) spectra were matched to human protein sequences after database searching and identified proteins were selected as potential biomarker candidates according to literature searches. Protein biomarker candidates were validated by Western blotting and immunohistochemistry to confirm the expression and the distribution of these proteins. The TMA of human Opisthorchis viverrini-associated CCA tissues was applied to provide a high-throughput validation of potential targets identified in the proteomics study.

3. Results

3.1. Protein expression patterns in CCA cell lines

In order to examine the differential protein profiles of CCA cell lines, three replicate maps for each cell type (M156, K100, M139, M213 and H69) were generated. The separated protein spots were visualised on 2-D gels by CBR-250 staining, which allows good reproducibility and protein spot quantification for comparative analysis. The 2-D gel containing the membrane proteins (Fig. 3) showed a total of 186 ± 20, 314 ± 93, 320 ± 17, 210 ± 35, and 270 ± 6 protein spots, for the M156, K100, M139, M213 and H69 cell lines, respectively. Using 2D gel replicates, identical and differential protein expression in four CCA cell lines subtracted from H69 were measured. In Table 1, the proteins corresponding to spots that were expressed only in CCA cell lines but not in H69 are shown, together with the MS identification parameters. In the tumor histology comparison, mitochondrial ATP synthase, keratin 8, aldo-keto reductase family 1, manganese superoxide dismutase (MnSOD) and heat shock protein 70 (Hsp70) were found only in K100 versus keratin 7 which was found only in M156. Splicing factor proline/glutamine-rich and calreticulin were expressed in M139 whereas the epithelial cell marker protein 1, enolase 1 and tubulin-beta were observed only in M213. We found the solute carrier family 25 protein (SLC25) was expressed in all CCA cell lines as were tubulin-alpha and keratin 18, except the latter was not expressed in M156. The Tu translation elongation factor and keratin 17 were found in the CCA cell lines except for K100. Glycerealdehyde-3 phosphate dehydrogenase, ANXA2, and keratin 19 were found in M213 and M139 compared with chaperonin-containing TCP1 found only in M139 and M156. Although SLC25 was found in all CCA cell lines, the aberrant protein expression of SLC25 has not been reported to be associated with any human tumors to date. In contrast, ANXA2 has been reported to play a key role in tumor invasion and metastasis in multiple neoplasms (Bao et al., 2009; Braden et al., 2009; Nedjadi et al 2009; Zhang et al., 2009) but has not yet been reported as being up-regulated in CCA. Accordingly, in this study, we selected ANXA2 for further verification using Western blots for membrane proteins from M139 and IHC for CCA tissues on the TMA (n = 301).

Fig 3.

Fig 3

Open in a new tab

Two-dimensional gel electrophoresis (2-DE) map of Opisthorchis viverrini-induced cholangiocarcinoma (CCA) cell lines displaying membrane protein profiles from M156 moderately differentiated CCA (A), K100 poorly differentiated CCA (B), M139 squamous cell carcinoma (C) and M213 adenosquamous cell carcinoma (D); using 7 cm pH 3–10 immobilized pH gradient (IPG) non-linear strips in the first dimension and 12.5% SDS-PAGE gel in the second dimension. The protein loading was 100 µg and the gel was stained with CBR-250.

Table 1.

Membrane proteins differentially expressed in Opisthorchis viverrini-induced cholangiocarcinoma (CCA) cell lines identified by MALDI-MS-TOF.

Accession No MS CO M/UM Description Mr pI Normal cell line
 CCA cell line
H69 M156 K100 M139 M213
gi|46249805 96 23 10/12 Solute carrier family 25 53.5 6.0 − + + + +
gi|109096498 230 61 17/16 Tubulin, alpha 53.5 4.9 − − + + +
gi|4557888 101 40 14/22 keratin 18 48.0 5.3 − − + + +
gi|4557701 454 73 35/15 keratin 17 48.4 5.0 − + − + +
gi|34147630 328 61 27/6 Tu translation elongation factor 53.5 7.3 − + − + +
gi|7669492 155 38 11/8 glyceraldehyde-3-phosphate
dehydrogenase		 36.2 8.6 − − − + +
gi|4757756 235 64 19/12 annexin A2 isoform 2 38.8 7.6 − − − + +
gi|24234699 336 63 26/9 keratin 19 44.1 5. 0 − − − + +
gi|5453603 129 31 13/15 chaperonin containing TCP1, subunit 2 57.8 6.0 − + − + −
gi|67782365 322 55 27/10 keratin 7 51.4 5.4 − + − − −
gi|89574029 102 66 27/18 mitochondrial ATP synthase 48.1 5.0 − − + − −
gi|24234688 128 52 31/9 heat shock 70 73.9 5.9 − − + − −
gi|4504919 364 57 34/10 keratin 8 53.7 5.5 − − + − −
gi|4502049 193 41 15/7 aldo-keto reductase family 1 36.2 6.5 − − + − −
gi|14488599 162 45 12/26 Manganese Superoxide Dismutase 22.2 6.9 − − + − −
gi|119627826 150 40 18/12 splicing factor proline/glutamine-rich 66.4 9.2 − − − + −
gi|62897681 172 41 18/26 calreticulin 47.1 4.3 − − − + −
gi|187302 319 77 23/26 epithelial cell marker protein 1 27.9 4.7 − − − − +
gi|4503571 156 66 24/4 enolase 1 47.5 7.0 − − − − +
gi|57209813 191 36 23/12 tubulin, beta polypeptide 48.1 4.7 − − − − +
Open in a new tab

The presence of a protein in the relevant study is denoted with a ‘+’ and the absence with a ‘-’.

MS, Mowse score (P < 0.05); CO, percentage of sequence coverage; M/UM, the number of matched peaks/unmatched peaks.

3.2. Western blotting and MS

Membrane proteins from M139 cells were separated by 2-DE and electro-blotted onto nitrocellulose membrane to be probed with commercially available ANXA2 antibody. Fig. 4A shows the corresponding CBR-250-stained gel. Fig. 4B shows the image obtained from analyzing M139 membrane proteins by 2-D immunoblotting with anti-ANXA2 antibody. Three different electrophoretic spots proved to be immunoreactive in the experimental window reported (pH 3–10, mass 10–200 kDa) showing approximately the same Mr, but different pI values. The presence of protein variants differing in charge through amino acid differences or post-translational modifications is the most straightforward explanation of these shifts (Pastorelli et al., 2007). Representative mass spectra and the sequences of the assigned peptides for ANXA2 isoform 2 are shown in Fig. 5.

Fig 4.

Fig 4

Open in a new tab

Two dimensional electropherogram of purified membrane protein from the M139 Opisthorchis viverrini-induced cholangiocarcinoma cell line (A). Two-dimensional immunoblotting with anti-Annexin A2 (ANXA2) antibody (B). Three immunoreactive spots are indicated (spot a: 38.8 kDa, pI 7.57; spot b: 38.8 kDa, pI 7.6; spot c: 38.8 kDa, pI 7.65).

Fig. 5.

Fig. 5

Open in a new tab

MALDI-TOF peptide mass fingerprint of the tryptic digest of annexin A2 (ANXA2). The MALDI-TOF mass spectrum of ANXA2, yielding 64% protein sequence coverage (A) and the matched peptide sequences were underlined (B).

3.3. Expression of ANXA2 and clinicopathological findings

IHC analyses were carried out on the TMAs (n = 301) to determine whether ANXA2 expression changes during cholangiocarcinogenesis. Hyperplastic bile duct (Fig. 6B) showed the same ANXA2 expression pattern as carcinogenic cells, in contrast with the loss of ANXA2 signal detected in the corresponding normal bile duct epithelium (Fig. 6A). ANXA2 staining was preferably membranous in location (Fig. 6D), although some cytoplasmic staining was observed (Fig. 6C). IHC demonstrated that ANXA2 was frequently expressed in CCA. Biopsies from 227 patients (76.2%) stained positively for ANXA2. Increased expression of ANXA2 did not show a preference for any histological subtypes of tumor or tumor size. However, statistical analyses showed that positive staining for ANXA2 significantly associated with poor prognosis. High expression of ANXA2 was found to be associated with lymphatic invasion (P = 0.014), metastasis (P = 0.026) and perineural invasion (P = 0.009) (Table 2).

Fig. 6.

Fig. 6

Open in a new tab

Immunohistochemical staining of Annexin A2 (ANXA2) in normal liver tissue (A), bile duct hyperplasia tissue (B) and cholangiocarcinoma (CCA) tissues (C and D). Annexin A2-positive cells were clustered within bile duct hyperplasia (B) and CCA tissues (C and D), but not detected or expressed at very low levels in stroma, normal liver and bile duct cells (A). Annexin A2 was preferably membranous (D) in location of CCA tissues, although some cytoplasmic staining (C) was observed. Immunoperoxidase staining, original magnification ×200 (A–D); scale bar = 50 µm.

Table 2.

Clinicopathological variables and the expression status of annexin A2 (ANXA2) in Opisthorchis viverrini-induced cholangiocarcinoma tissues.

Variables Annexin A2
 P
−ve +ve
Age(yr)
  ≤ 56 50 114 0.001
  > 56 19 111
Gender
  Male 46 156 NS
  Female 24 70
Histotype group
  Less diff. 42 130 NS
  Well diff. 19 76
Gross type
  Mass forming 36 108 NS
  Periductal infiltrating 11 49
  Intraductal 2 7
Tumor size
  ≤ 5 cm 28 87 NS
  > 5 cm 27 59
Vascular invasion
  Absent 34 100 NS
  Present 32 118
Lymphatic invasion
  Absent 25 49 0.014
  Present 41 167
Perineural invasion
  Absent 49 121 0.009
  Present 17 94
Metastasis
  Absent 36 92 0.026
  Present 20 102
Open in a new tab

When the sum of subset numbers does not match patient totals, data were missing or unavailable. NS, not significant

3.4. Expression of ANXA2 and cumulative survival

The median length of survival for patients with high ANXA2 expression was 27.42 weeks (95% Confidence Interval (95% CI), 22.26–32.58 weeks) whereas it was 42.86 weeks (95% CI, 25.92–59.80 weeks) for patients with low ANXA2 expression. Survival curves for the patients were generated based on the ANXA2 staining categories (Fig. 7). Diminished survival was seen in the cases with high expression of ANXA2 (log-rank test, P = 0.011).

Fig. 7.

Fig. 7

Open in a new tab

Survival curves using the Kaplan-Meier method. Significant unfavorable prognosis of cholangiocarcinoma (CCA) was high compared with low expression for annexin A2 (ANXA2) (P = 0.011). Annexin A2 (−) = low expression, annexin A2 (+) = high expression.

Ten potential prognostic factors were analyzed for their risk in the survival of O. viverrini-associated CCA (age, gender, histotype, gross type, tumor size, vascular invasion, lymphatic invasion, perineural invasion, metastasis, ANXA2 expression). In univariate analyses (log-rank test), vascular invasion (P = 0.003), lymphatic invasion (P = 0.001), metastasis (P = 0.001) and ANXA2 expression (P = 0.011) were found to be associated with survival time after surgery. Multivariate analysis (Cox-regression) was performed using the above four significant variables; lymphatic invasion (P = 0.001), metastasis (P = 0.003) and ANXA2 expression (P = 0.046) were identified as independent prognostic factors in O. viverrini-associated CCA (Table 3).

Table 3.

Risk factors for overall survival in Opisthorchis viverrini-associated cholangiocarcinoma.

Risk factors Overall survival
Univariate analysis
(Log-rank)
 Multivariate analysis
(Cox regression)
Median time, wk P Relative risk 95%CI P
Vascular invasion
  Absent 38.14 ± 3.06 0.003 0.78 0.61–1.00 NS
  Present 25.42 ± 1.90 1
Lymphatic invasion
  Absent 44.00 ± 4.82 0.001 0.59 0.43–0.81 0.001
  Present 27.28 ± 1.95 1
ANXA2
  Low expression 42.86 ± 8.64 0.011 0.71 0.51–0.99 0.046
  High expression 27.42 ± 2.63 1
Metastasis
  Absent 38.14 ± 4.57 0.001 0.65 0.49–0.86 0.003
  Present 26.57 ± 2.65
Open in a new tab

CI, confident interval; NS, not significant

4. Discussion

The complex biological fundamentals of CCA are still poorly elucidated, despite intensive studies. Several mechanisms by which O. viverrini infection may enhance cholangiocarcinogenesis have been proposed (Sripa and Pairojkul, 2008). As in other epithelial neoplasms, it seems to develop through multiple processes involving DNA damage, deregulation of epithelial differentiation and abnormal cell proliferation (Sripa and Pairojkul, 2008). With less apoptosis of the infected biliary epithelium, after several series of replication, genetic alterations may become fixed which can lead to neoplasm transformation (Sripa et al., 2007). Identifying the protein alterations associated with these events could be an important step towards understanding the mechanisms of cholangiocarcinogenesis and could facilitate the development of new tools for diagnosis, treatment and prevention of CCA.

Annexins are characterized by their capacity to bind to phospholipids in the presence of calcium ions and their susceptibility to phosphorylation and dephosphorylation (Rodrigo Tapia et al., 2007). Belonging to subfamily A of annexins of vertebrates, human annexins are further classified as annexin A1-A11 and A13 (Rodrigo Tapia et al., 2007). As show in Fig. 8, ANXA2 consists of a variable N-terminal “tail” domain that imparts specialized features to the molecule and a conserved C-terminal “core” domain that imparts phospholipid binding capacity (Brownstein et al., 2001; Deora et al., 2004). ANXA2 has been shown to exist as a monomer, a heterodimer or a heterotetramer with S100A10 (light chain of 11 kDa, or p11) (Waisman, 1995). ANXA2 is located in the cell cytoplasm as a monomer (heavy chain of 36 kDa, or p36) or in a complex with a member of the S100A10 (Rodrigo Tapia et al., 2007). Heterotetramer formation between two copies of ANXA2 and two copies of S100A10 results in enhanced membrane phospholipid binding affinity (Erikson and Erikson, 1980). Recent evidence indicates that ANXA2 is a constitutive regulator of the fibrinolytic process when located on the cell surface. ANXA2 serves as a receptor or binding protein for proteases (cathepsin B, plasminogen and tissue plasminogen activator or tPA) and proteins in the extracellular matrix (collagen and tenascin C) (Rodrigo Tapia et al., 2007). A relationship between proteases and extracellular matrix proteins through ANXA2 has been proposed whereby ANXA2 may facilitate the reorganisation of the extracellular matrix in physiological and pathological processes such as tumoral invasion (Mai et al., 2000). Based on this reasoning, the over-expression of ANXA2 has been verified in colorectal (Emoto et al., 2001b) and gastric (Emoto et al., 2001a) carcinomas, and correlates with invasiveness and poor prognosis.

Fig. 8.

Fig. 8

Open in a new tab

Structure of annexin A2 (ANXA2), shown as a heterotetramer with p11 (S100A10, light chain of 11 kDa). Similar to other family members, ANXA2 consist of a conserved “core” domain and a highly variable amino terminal tail domain. The four annexin repeats found within the core domain of the ANXA2 were represented as 1–4. The phosphorylation site of tyrosine and serine is presented within the tail domain. Figure adapted from Brownstein et al., 2001.

The expression of ANXA2 had already been described in tissues and cell lines of Korean ICC patients using expressed sequence tags (Wang et al., 2006), but there was a limited sample size which precluded IHC investigations (12 human ICC tissues and four normal liver tissues). Moreover, there are no data available on the expression of ANXA2 in O. viverrini-associated CCA which is commonly found in the northeast of Thailand.

In this study, we applied a proteomics-based approach to find differentially expressed tumor proteins in CCA cell lines. However, a drawback in using cell lines is that the cells are not representative of the tissue of origin because they have been maintained and propagated in vitro for extended periods (Hay, 1988). Furthermore, each cell line was established from a few cells of a single patient biopsy, so cells may not accurately represent the tumor characteristics (Hay, 1988). Accordingly, validation of potential candidate tumor markers in patient biopsies is essential. ANXA2 was shown to be up-regulated in a proteomics study of cell lines among many other markers. Again, ANXA2 is known to play an important role in tumor invasion and metastasis in other cancers (Emoto et al., 2001a; Emoto et al., 2001b). Accordingly, ANXA2 was verified in human subjects by probing, using a commercial anti-mouse monoclonal antibody in the TMA of patients with CCA (301 diagnosed cases), where it was found to associate with one of several tumor progression stages as reflected by lymphatic invasion (P = 0.014) and metastasis (P = 0.026). In addition, the over-expression of ANXA2 found to be associated with decreasing perineural invasion in CCA tissues has been linked to tumoral suppression and the inhibition of cell migration as previously studied in other cancers (Liu et al., 2003; Rodrigo Tapia et al., 2007). A Cox-regression model was used in multifactorial analysis (age, gender, histotype, gross type, tumor size, vascular invasion, lymphatic invasion, perineural invasion, metastasis, ANXA2 expression and survival time after surgery), indicating ANXA2, vascular invasion, lymphatic invasion and metastasis had an association with survival time after surgery. As shown in Table 3, patients with low expression of ANXA2 had longer survival times after surgery compared with the patients with high expression of ANXA2, indicating that over-expression of ANXA2 may reflect tumor progression of O. viverrini-associated CCA. Regarding vascular invasion, lymphatic invasion and metastasis in our multifactorial analysis, it has been previously reported that the survival time after surgery for the group with the presence of tumor invasion and metastasis was shorter than the group with absence of tumor invasion and metastasis in CCA (Uttaravichien et al., 1999).

In conclusion, we have described, to our knowledge for the first time, that ANXA2 is expressed both in the cytoplasm and cell membrane of hyperplastic epithelia and CCA but not in normal cholangiocytes and hepatocytes. Our results indicate that up-regulated ANXA2 may serve as a prognostic marker for invasion, metastasis and survival in O. viverrini-associated CCA. We suggest that ANXA2 IHC may be applied for diagnosis of CCA and prediction of patient outcome. Proteomic studies of tumor tissue and serum from patients with O. viverrini-associated CCA are of particular interest to aid the discovery of tumor markers and are currently underway in our laboratories.

Acknowledgements

This work was supported in part by the ICIDR-NIAID, NIH, USA (award number UO1AI065871). Ponlapat Yonglitthipagon is a Royal Golden Jubilee PhD scholar (grant PHD/0194/2548) through the laboratory of Dr. Banchob Sripa.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bao H, Jiang M, Zhu M, Sheng F, Ruan J, Ruan C. Overexpression of Annexin II affects the proliferation, apoptosis, invasion and production of proangiogenic factors in multiple myeloma. Int J Hematol. 2009;90:177–185. doi: 10.1007/s12185-009-0356-8. [DOI] [PubMed] [Google Scholar]
  2. Blechacz BR, Gores GJ. Cholangiocarcinoma. Clin Liver Dis. 2008;12:131–150. doi: 10.1016/j.cld.2007.11.003. ix. [DOI] [PubMed] [Google Scholar]
  3. Bouvard V, Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Benbrahim-Tallaa L, Guha N, Freeman C, Galichet L, Cogliano V. A review of human carcinogens--Part B: biological agents. Lancet Oncol. 2009;10:321–322. doi: 10.1016/s1470-2045(09)70096-8. [DOI] [PubMed] [Google Scholar]
  4. Boyd RS, Dyer MJ, Cain K. Proteomic analysis of cell surface membrane proteins in leukemic cells. Methods Mol Biol. 2007;370:135–146. doi: 10.1007/978-1-59745-353-0_11. [DOI] [PubMed] [Google Scholar]
  5. Braden AR, Kafka MT, Cunningham L, Jones H, Vishwanatha JK. Polymeric nanoparticles for sustained down-regulation of annexin A2 inhibit prostate tumor growth. J Nanosci Nanotechnol. 2009;9:2856–2865. doi: 10.1166/jnn.2009.028. [DOI] [PubMed] [Google Scholar]
  6. Brownstein C, Falcone DJ, Jacovina A, Hajjar KA. A mediator of cell surface-specific plasmin generation. Ann N Y Acad Sci. 2001;947:143–155. [PubMed] [Google Scholar]
  7. Deora AB, Kreitzer G, Jacovina AT, Hajjar KA. An annexin 2 phosphorylation switch mediates p11-dependent translocation of annexin 2 to the cell surface. J Biol Chem. 2004;279:43411–43418. doi: 10.1074/jbc.M408078200. [DOI] [PubMed] [Google Scholar]
  8. Emoto K, Sawada H, Yamada Y, Fujimoto H, Takahama Y, Ueno M, Takayama T, Uchida H, Kamada K, Naito A, Hirao S, Nakajima Y. Annexin II overexpression is correlated with poor prognosis in human gastric carcinoma. Anticancer Res. 2001a;21:1339–1345. [PubMed] [Google Scholar]
  9. Emoto K, Yamada Y, Sawada H, Fujimoto H, Ueno M, Takayama T, Kamada K, Naito A, Hirao S, Nakajima Y. Annexin II overexpression correlates with stromal tenascin-C overexpression: a prognostic marker in colorectal carcinoma. Cancer. 2001b;92:1419–1426. doi: 10.1002/1097-0142(20010915)92:6<1419::aid-cncr1465>3.0.co;2-j. [DOI] [PubMed] [Google Scholar]
  10. Erikson E, Erikson RL. Identification of a cellular protein substrate phosphorylated by the avian sarcoma virus-transforming gene product. Cell. 1980;21:829–836. doi: 10.1016/0092-8674(80)90446-8. [DOI] [PubMed] [Google Scholar]
  11. Fedor HL, De Marzo AM. Practical methods for tissue microarray construction. Methods Mol Med. 2005;103:89–101. doi: 10.1385/1-59259-780-7:089. [DOI] [PubMed] [Google Scholar]
  12. Grubman SA, Perrone RD, Lee DW, Murray SL, Rogers LC, Wolkoff LI, Mulberg AE, Cherington V, Jefferson DM. Regulation of intracellular pH by immortalized human intrahepatic biliary epithelial cell lines. Am J Physiol. 1994;266:G1060–1070. doi: 10.1152/ajpgi.1994.266.6.G1060. [DOI] [PubMed] [Google Scholar]
  13. Hay MM, Chen TR, McClintock P, Reid Y. American Type Culture CollectionCatalogue of Cell Lines and Hybridomas. 6th ed. Rockville, Md: American Type Culture Collection; 1988. [Google Scholar]
  14. IARC. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Lyon, 24 February–3 March 2009. Schistosoma haematobium, liver flukes and Heliobacter pylori . IARC Monogr Eval Carcinog Risks Hum. 2009;100 in press. [Google Scholar]
  15. Ji NY, Park MY, Kang YH, Lee CI, Kim DG, Yeom YI, Jang YJ, Myung PK, Kim JW, Lee HG, Kim JW, Lee K, Song EY. Evaluation of annexin II as a potential serum marker for hepatocellular carcinoma using a developed sandwich ELISA method. Int J Mol Med. 2009;24:765–771. doi: 10.3892/ijmm_00000290. [DOI] [PubMed] [Google Scholar]
  16. Jongsuksuntigul P, Imsomboon T. Opisthorchiasis control in Thailand. Acta Trop. 2003;88:229–232. doi: 10.1016/j.actatropica.2003.01.002. [DOI] [PubMed] [Google Scholar]
  17. Kischel P, Guillonneau F, Dumont B, Bellahcene A, Stresing V, Clezardin P, De Pauw EA, Castronovo V. Cell membrane proteomic analysis identifies proteins differentially expressed in osteotropic human breast cancer cells. Neoplasia. 2008;10:1014–1020. doi: 10.1593/neo.08570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Liu JW, Shen JJ, Tanzillo-Swarts A, Bhatia B, Maldonado CM, Person MD, Lau SS, Tang DG. Annexin II expression is reduced or lost in prostate cancer cells and its re-expression inhibits prostate cancer cell migration. Oncogene. 2003;22:1475–1485. doi: 10.1038/sj.onc.1206196. [DOI] [PubMed] [Google Scholar]
  19. Mai J, Waisman DM, Sloane BF. Cell surface complex of cathepsin B/annexin II tetramer in malignant progression. Biochim Biophys Acta. 2000;1477:215–230. doi: 10.1016/s0167-4838(99)00274-5. [DOI] [PubMed] [Google Scholar]
  20. Mohammad HS, Kurokohchi K, Yoneyama H, Tokuda M, Morishita A, Jian G, Shi L, Murota M, Tani J, Kato K, Miyoshi H, Deguchi A, Himoto T, Usuki H, Wakabayashi H, Izuishi K, Suzuki Y, Iwama H, Deguchi K, Uchida N, Sabet EA, Arafa UA, Hassan AT, El-Sayed AA, Masaki T. Annexin A2 expression and phosphorylation are up-regulated in hepatocellular carcinoma. Int J Oncol. 2008;33:1157–1163. [PubMed] [Google Scholar]
  21. Nakanuma Y, Sripa B, Vatanasapt V, Leong A, Ponchon T, Ishak K. Intrahepatic cholangiocarcinoma. In: Hamilton SR, Aaltonen LA, editors. WHO Classification of Tumors, Pathology and Genetics. Tumours of the Digestive System. Lyon: IARC Press; 2000. pp. 173–180. [Google Scholar]
  22. Nedjadi T, Kitteringham N, Campbell F, Jenkins RE, Park BK, Navarro P, Ashcroft F, Tepikin A, Neoptolemos JP, Costello E. S100A6 binds to annexin 2 in pancreatic cancer cells and promotes pancreatic cancer cell motility. Br J Cancer. 2009;101:1145–1154. doi: 10.1038/sj.bjc.6605289. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Olnes MJ, Erlich R. A review and update on cholangiocarcinoma. Oncology. 2004;66:167–179. doi: 10.1159/000077991. [DOI] [PubMed] [Google Scholar]
  24. Pastorelli R, Saletta F, Carpi D, Campagna R, dell’Osta C, Schiarea S, Vineis P, Airoldi L, Matullo G. Proteome characterization of a human urothelial cell line resistant to the bladder carcinogen 4-aminobiphenyl. Proteome Sci. 2007;5:6. doi: 10.1186/1477-5956-5-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rodrigo Tapia JP, Pena Alonso E, Garcia-Pedrero JM, Florentino Fresno M, Suarez Nieto C, Owen Morgan R, Fernandez MP. [Annexin A2 expression in head and neck squamous cell carcinoma] Acta Otorrinolaringol Esp. 2007;58:257–262. [PubMed] [Google Scholar]
  26. Ruan Y, Wan M. An optimized procedure for solubilization, reduction, and transfer of human breast cancer membrane-enriched fraction by 2-DE. Electrophoresis. 2007;28:3333–3340. doi: 10.1002/elps.200700035. [DOI] [PubMed] [Google Scholar]
  27. Sardana G, Marshall J, Diamandis EP. Discovery of candidate tumor markers for prostate cancer via proteomic analysis of cell culture-conditioned medium. Clin Chem. 2007;53:429–437. doi: 10.1373/clinchem.2006.077370. [DOI] [PubMed] [Google Scholar]
  28. Sriamporn S, Pisani P, Pipitgool V, Suwanrungruang K, Kamsaard S, Parkin DM. Opisthorchis viverrini infection and incidence of cholangiocarcinoma in Khon Kaen, Northeast Thailand. Trop Med Int Health. 2004. Prevalence of;9:588–594. doi: 10.1111/j.1365-3156.2004.01234.x. [DOI] [PubMed] [Google Scholar]
  29. Sriamporn S, Parkin DM, Pisani P, Vatanasapt V, Suwanrungruang K, Kamsa-ard P, Pengsaa P, Kritpetcharat O, Pipitgool V, Vatanasapt P. A prospective study of diet, lifestyle, and genetic factors and the risk of cancer in Khon Kaen Province, northeast Thailand: description of the cohort. Asian Pac J Cancer Prev. 2005;6:295–303. [PubMed] [Google Scholar]
  30. Sripa B, Leungwattanawanit S, Nitta T, Wongkham C, Bhudhisawasdi V, Puapairoj A, Sripa C, Miwa M. Establishment and characterization of an opisthorchiasis-associated cholangiocarcinoma cell line (KKU-100) World J Gastroenterol. 2005;11:3392–3397. doi: 10.3748/wjg.v11.i22.3392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sripa B, Kaewkes S, Sithithaworn P, Mairiang E, Laha T, Smout M, Pairojkul C, Bhudhisawasdi V, Tesana S, Thinkamrop B, Bethony JM, Loukas A, Brindley PJ. Liver fluke induces cholangiocarcinoma. PLoS Med. 2007;4:e201. doi: 10.1371/journal.pmed.0040201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sripa B, Pairojkul C. Cholangiocarcinoma: lessons from Thailand. Curr Opin. Gastroenterol. 2008;24:349–356. doi: 10.1097/MOG.0b013e3282fbf9b3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Uttaravichien T, Bhudhisawasdi V, Pairojkul C, Pugkhem A. Intrahepatic cholangiocarcinoma in Thailand. J Hepatobiliary Pancreat Surg. 1999;6:128–135. doi: 10.1007/s005340050095. [DOI] [PubMed] [Google Scholar]
  34. van Diest PJ, van Dam P, Henzen-Logmans SC, Berns E, van der Burg ME, Green J, Vergote I. A scoring system for immunohistochemical staining: consensus report of the task force for basic research of the EORTC-GCCG. European Organization for Research and Treatment of Cancer-Gynaecological Cancer Cooperative Group. J Clin Pathol. 1997;50:801–804. doi: 10.1136/jcp.50.10.801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Waisman DM. Annexin II tetramer: structure and function. Mol Cell Biochem. 1995;149–150:301–322. doi: 10.1007/BF01076592. [DOI] [PubMed] [Google Scholar]
  36. Wang AG, Yoon SY, Oh JH, Jeon YJ, Kim M, Kim JM, Byun SS, Yang JO, Kim JH, Kim DG, Yeom YI, Yoo HS, Kim YS, Kim NS. Identification of intrahepatic cholangiocarcinoma related genes by comparison with normal liver tissues using expressed sequence tags. Biochem Biophys Res Commun. 2006;345:1022–1032. doi: 10.1016/j.bbrc.2006.04.175. [DOI] [PubMed] [Google Scholar]
  37. Watanapa P. Cholangiocarcinoma in patients with opisthorchiasis. Br J Surg. 1996;83:1062–1064. doi: 10.1002/bjs.1800830809. [DOI] [PubMed] [Google Scholar]
  38. Zhang F, Zhang L, Zhang B, Wei X, Yang Y, Qi RZ, Ying G, Zhang N, Niu R. Anxa2 plays a critical role in enhanced invasiveness of the multidrug resistant human breast cancer cells. J Proteome Res. 2009;8:5041–5047. doi: 10.1021/pr900461c. [DOI] [PubMed] [Google Scholar]
  39. Zhuang PY, Zhang JB, Zhu XD, Zhang W, Wu WZ, Tan YS, Hou J, Tang ZY, Qin LX, Sun HC. Two pathologic types of hepatocellular carcinoma with lymph node metastasis with distinct prognosis on the basis of CK19 expression in tumor. Cancer. 2008;112:2740–2748. doi: 10.1002/cncr.23488. [DOI] [PubMed] [Google Scholar]

RESOURCES