Reg IV expression and clinicopathologic features of gallbladder carcinoma

Human Pathology (2009) 40, 1686–1692

www.elsevier.com/locate/humpath

Original contribution

Reg IV expression and clinicopathologic features of gallbladder carcinoma Hidehiko Tamura MD, PhD a , Masayuki Ohtsuka MD, PhD a,⁎, Mitsutsune Washiro MD, PhD a , Fumio Kimura MD, PhD a , Hiroaki Shimizu MD, PhD a , Hiroyuki Yoshidome MD, PhD a , Atsushi Kato MD, PhD a , Naohiko Seki PhD b , Masaru Miyazaki MD, PhD a a

Department of General Surgery, Graduate School of Medicine, Chiba University, Chiba 260-8670, Japan Department of Functional Genomics, Graduate School of Medicine, Chiba University, Chiba 260-8670, Japan

b

Received 19 January 2009; revised 2 June 2009; accepted 11 June 2009

Keywords: Reg IV; Gallbladder cancer; Intestinal metaplasia; Cdx2

Summary Regenerating islet-derived family, member 4 (Reg IV) has been shown to be associated with colorectal carcinogenesis and gastric carcinogenesis through intestinal metaplasia. In this study, we examined Reg IV expression in the gallbladder and gallbladder carcinoma, and measured Reg IV levels in sera from patients with gallbladder carcinoma. Quantitative reverse transcription–polymerase chain reaction revealed that high Reg IV levels were identified in 17 of 31 gallbladder carcinomas, whereas there was no apparent amplification in normal gallbladders. Immunohistochemically, although only a small part of the epithelium with intestinal metaplasia in 2 of 4 cases with adenomyomatosis showed Reg IV expression, Reg IV was negative in all cases with normal gallbladder (n = 15) and cholelithiasis (n = 13). In contrast, 34 (56%) of 61 gallbladder carcinomas were positive. Expression was more frequently observed in well to moderately differentiated than in poorly differentiated adenocarcinomas and significantly correlated with expression of caudal-related homeobox transcription factor (a candidate for involvement in the induction of intestinal metaplasia). Multivariate analysis revealed negative Reg IV expression, as well as hepatic parenchymal invasion, to be independently associated with a poor prognosis in patients with advanced gallbladder carcinoma. Before surgical resection, 4 (33%) of 12 patients with gallbladder carcinoma had high serum Reg IV levels, whereas Reg IV was never elevated in 12 patients with benign diseases. The serum levels of Reg IV decreased after surgical resection of the tumors. These results suggest that Reg IV is involved in gallbladder carcinoma carcinogenesis through intestinal metaplasia and is associated with relatively favorable prognosis in patients after surgery. The serum level of Reg IV may be of use or indicative of neoplasia. © 2009 Elsevier Inc. All rights reserved.

1. Introduction

Abbreviations: Cdx2, caudal-related homeobox transcription factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GBC, gallbladder carcinoma; Reg IV, regenerating islet-derived family, member 4. ⁎ Corresponding author. E-mail address: [email protected] (M. Ohtsuka). 0046-8177/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2009.06.001

The regenerating islet-derived family, member 4 (Reg IV) gene was originally identified by high-throughput sequencing of a complementary DNA library derived from patients with inflammatory bowel disease [1] and encodes a secreted Reg IV protein. Reg IV is known to be expressed in various normal tissues such as the stomach,

Reg IV expression in the gallbladder and gallbladder carcinoma colon, small intestine, and pancreas [2,3]. In cancers, although lung and breast cancers do not express Reg IV, gastric, colorectal, pancreatic, and prostate cancers have been reported to express Reg IV, and the levels of expression in these cancers are much higher than in normal tissues [2-4]. The biological function of Reg IV in cancers, however, has been poorly understood, although several possible functions have been reported. Among these, several studies have shown that Reg IV may be involved in early carcinogenesis in certain cancers. Zhang et al [5] reported that overexpression of Reg IV may be an early event in colorectal carcinogenesis, on the basis of their observations in which Reg IV mRNA expression was upregulated in colorectal adenomas and adenocarcinomas when compared to paired normal mucosa, and the staining of Reg IV by in situ hybridization was stronger in adenomas with more severe dysplasia. Oue et al [2] showed that Reg IV protein was expressed in the goblet cells of intestinal metaplasia of the stomach and gobletlike cell vesicles of gastric carcinoma, suggesting the association of Reg IV protein with intestinal differentiation of the stomach and tumorigenesis of intestinal-type gastric carcinoma. Gallbladder carcinoma (GBC) is a highly malignant neoplasm. Despite the improvement of diagnostic modalities, most cases with GBC are diagnosed at advanced stages, resulting in poor prognosis even after surgical resection [6-8]. Although the mechanisms of carcinogenesis of GBC have not been fully evaluated, dysplasia is considered to be a precursor of invasive GBC, because of their close topographical relationship [9,10]. In addition, several reports have shown that the most common epithelial change associated with dysplasia is intestinal metaplasia [911], suggesting that dysplasia arises from metaplasia, which is possibly attributable to chronic inflammation in the mucosa of the gallbladder [12,13]. Furthermore, Mukhopadhyay and Landas [14] proposed a sequence of gastric (antral) metaplasia–intestinal metaplasia–dysplasia, as evidenced by an age gradient from antral metaplasia to dysplasia, and significant statistical associations between antral metaplasia and intestinal metaplasia and intestinal metaplasia and dysplasia. This sequence is similar to that in the gastric carcinogenic pathway: chronic inflammation and regeneration lead to intestinal metaplasia in the gastric mucosa, followed by dysplasia, and, finally, intestinal-type gastric carcinoma [15]. These observations led us to the idea that Reg IV may also play an important role in gallbladder carcinogenesis through intestinal metaplasia. On the other hand, it has been reported that intestinal metaplasia and caudal-related homeobox transcription factor (Cdx2) in the gallbladder are related [12,16]. Cdx2 plays a crucial role in the regulation of intestinal epithelial differentiation and the maintenance of its phenotype [17]. Besides the expression in the fetal and adult epithelial nuclei of normal intestines, Cdx2 is ectopically expressed

1687

in the nuclei of the epithelium with intestinal metaplasia in some organs and some intestinal-type neoplasms. Although in intraductal papillary mucinous neoplasms of the pancreas Cdx2 expression is shown to be significantly correlated with Reg IV expression [18], the relationship between Reg IV and Cdx2 expressions remains unknown in GBC. In this study, we investigated the expression of Reg IV mRNA and its protein in the gallbladder and GBC using molecular and immunohistochemical techniques. The relationship between Reg IV and Cdx2 protein expressions in GBC was also examined. In addition, because Reg IV is a secreted protein, we measured Reg IV levels in sera from patients with GBC by ELISA to assess the possibility of Reg IV as a serum biomarker for GBC.

2. Materials and methods 2.1. Tissue samples A total of 31 fresh specimens from 31 patients with GBC, who underwent surgical resection at the Chiba University Hospital, were made available for genetic analysis with fully informed consent (average age, 64.5 years; male/female, 8/23). Tumor stage of these patients was stage IA in 1 patient, IB in 5, IIA in 4, IIB in 15, III in 3, and IV in 3, according to the pathologic tumor node metastasis classification [19]. Samples were snap frozen in liquid nitrogen and stored at −80°C until examined. In addition, gallbladder epithelium was obtained from 10 patients undergoing hepatic resection with cholecystectomy as a result of liver metastasis of colorectal cancer. These samples showing no definite histologic abnormalities were used as “normal gallbladder” samples. For immunohistochemical analysis, we used 61 formalinfixed, paraffin-embedded tissues obtained from 61 patients with GBC undergoing surgical resection with informed consent. Of the 61 patients, 9 had early GBC, which was limited to the mucosa (n = 6) or the mucosa and the muscularis propria (n = 3), and 52 had advanced GBC, which extended into the adventitia (n = 23) or beyond the adventitia (n = 29). In addition, 13 gallbladder tissue samples obtained from patients with cholelithiasis and 4 from patients with adenomyomatosis, as well as 15 normal gallbladders, were also used for immunohistochemistry. For both analyses, the cases with sclerosing cholangitis were not included.

2.2. Serum samples Among the 61 cases with GBC used for immunohistochemical analysis, preoperative and postoperative (at least 4 weeks after surgical resection) serum samples were available for ELISA with informed consent from 12

1688 cases, including 3 early GBC cases. Serum samples from 12 patients with cholelithiasis or adenomyomatosis were also collected. Control serum samples were obtained from 20 healthy volunteers.

2.3. Clinicopathologic data As variables possibly affecting prognosis, age and sex of patients, hepatic parenchymal invasion, bile duct invasion, cancer differentiation, invasion to the stomach or intestines, vascular invasion, neural invasion, lymph node involvement, and presence of cancer cells in the resection margin were recorded. Survival data (median length of follow-up, 57.9 months) were available for all patients.

2.4. Quantitative reverse transcription–polymerase chain reaction The cDNA template for reverse transcription–polymerase chain reaction (RT-PCR) was synthesized from the extracted total RNA (1-3 μg), using SuperScript II reverse transcriptase and an oligo(dT) primer (Invitrogen, Carlsbad, CA). PCR amplification was performed with LightCycler (Roche Diagnostics, Mannheim, Germany) in accordance with the manufacturer's instructions, using double stranded–specific dye SYBR Green I (Roche Diagnostics). The primer sequences for Reg IV were as follows: 5′-GGCAGCAGTGGCAGTGGATT-3′ and 5′-GGCGCTTGTTGCATTCGTTG-3′. PCR conditions consisted of an initial denaturation at 95 °C for 10 min, followed by 35 to 42 cycles of denaturation at 95 °C for 10 seconds, annealing at 62 °C for 10 seconds, and extension at 72 °C for 8 to 19 seconds. Expression of the housekeeping gene (ie, glyceraldehyde-3-phosphate dehydrogenase gene) was determined as an endogenous reference for each sample to normalize the amount of cDNA. To confirm specific amplification from Reg IV gene, melting curve analysis and electrophoretic analysis of the PCR products were performed.

2.5. Immunohistochemistry Standard streptavidin-biotin complex immunohistochemistry was performed with diaminobenzidine as a chromogen, and a light counterstain with hematoxylin. The primary antibodies used for this study were as follows: goat antihuman Reg IV polyclonal antibody (1:25, R&D Systems, Abingdon, UK) and mouse antihuman Cdx2 monoclonal antibody (1:20, Bio Genex, San Ramon, CA). Antigens were retrieved from deparaffinized and rehydrated tissues by autoclave treatment (121 °C, 15 minutes) in citrate buffer (pH 6.0) followed by cooling at room temperature for 30 minutes. Nonimmunized mouse immunoglobulin was treated instead of the primary antibody as a negative control. Normal pancreatic tissue served as positive controls for Reg IV and Cdx2, which were expressed in islet cells [2] and in the epithelium of the

H. Tamura et al. small pancreatic duct [20], respectively. For each case, 2 to 5 sections were examined. The staining of cytoplasm including the secretory vesicles in the mucosal epithelium of the gallbladder and cancer cells was evaluated for Reg IV, which was not detected in other cells. The percentage of the areas that showed Reg IV expression were graded as follows: negative, 0; less than 10%, 1+; 10% to 29%, 2+; 30% to 49%, 3+; greater than 50%, 4+. For Cdx2, only nuclear staining was regarded as expression [20], although cytoplasmic background staining was sometimes encountered. For both antibodies, the result of immunostaining was considered to be positive if more than 10% of the areas of interest were stained. When fewer than 10% of the areas were stained, the immunostaining was considered negative. This criterion is based on several previous studies [18,21-23].

2.6. Enzyme-linked immunosorbent assay Serum levels of Reg IV were measured by standard sandwich ELISA according to the description by Mitani et al [21] with a slight modification. Briefly, 96-well polystyrene microtiter plates (R&D Systems) were precoated with mouse monoclonal anti–Reg IV antibody (R&D Systems) by overnight incubation of 50 μL/125 ng per well of antibody diluted in phosphate buffered saline (pH 7.4) at room temperature. After blocking with 4% skim milk for 1 hour, 50 μL of recombinant Reg IV standard (Bio Vendor, Modrice, Czech Republic) or sample was incubated on the plate for 2 hours at room temperature. Biotinylated goat polyclonal antiReg IV antibody (R&D Systems) in ELISA assay buffer (1% bovine serum albumin, Tris buffer [pH 7.4], 0.05% normal goat serum) was used as secondary antibody and incubated for 2 hours with shaking at room temperature before addition of horse radish peroxidase–conjugated streptavidin (R&D Systems) diluted 200-fold in diluent containing 1% bovine serum albumin and Tris buffer (pH 7.4). After 20 minutes of incubation at room temperature, color was developed with the addition of TMB substrate solution (R&D Systems), followed by incubation for 20 minutes. Plates were analyzed with a spectrophotometer at 450 nm, using the Sunrise Classic microplate reader (Tecan, Mannedorf, Switzerland). Wells were washed 3 times in ELISA-washing buffer (R&D Systems) between each step of the ELISA. All samples were tested in triplicate.

2.7. Statistics Data were analyzed using the Fisher probability test or the χ2 test, to evaluate the significance of differences. Cumulative survival was calculated by the Kaplan-Meier method, and differences in univariate survival were examined using the log rank test. The Cox proportional hazards model was used for multivariate regression analysis. P b .05 was considered statistically significant.

Reg IV expression in the gallbladder and gallbladder carcinoma

1689

3. Results 3.1. Expression of Reg IV mRNA in GBC Quantitative RT-PCR revealed that normal gallbladders lacked an apparent amplification for Reg IV. In contrast, high levels of Reg IV (more than the mean level + 3 SD in normal gallbladders) were identified in 17 (55%) of 31 GBC tissues (Fig. 1).

3.2. Immunohistochemical analysis of Reg IV in gallbladder Reg IVexpression in the mucosal epithelium was negative in all 32 cases of normal gallbladder, cholelithiasis, and adenomyomatosis. Only a very small part of the epithelium (b10%) with intestinal metaplasia was found in 2 of 4 cases with adenomyomatosis, showing Reg IV expression in goblet cell vesicles (Fig. 2). In contrast, antral metaplasia, which was seen in 10 of 17 cases with cholelithiasis and adenomyomatosis, did not show Reg IV expression. Of 61 GBCs, 34 (56%) showed positive expression (≥2+) in the mucosa of Reg IV and its expression was detected focally (1+) in 11 (18%). Among 34 cases with positive Reg IV expression, 16 were graded as 2+, 10 as 3+, and 8 as 4+. Reg IV was detected in goblet-like cell vesicles or in the cytoplasm even without secretory vesicles of cancer cells (Fig. 3). When nonneoplasmic epithelial lesions could be evaluated in the surrounding mucosa of the tumoral lesion, Reg IV expression was also found in goblet cell vesicles of the dysplastic lesion with intestinal metaplasia and mucosal epithelium with intestinal metaplasia, but not in the epithelium with antral metaplasia. Positive Reg IV expression was observed even in early GBC, and there was no relationship between the frequency of Reg IV expression and the depth of tumor invasion. Well to moderately differentiated tubular-type adenocarcinoma showed positive Reg

Fig. 2 Immunohistochemical staining of the epithelium of intestinal metaplasia accompanied by adenomyomatosis of the gallbladder showing Reg IV expression in goblet-like cell vesicles. A, Hematoxylin and eosin staining; B, immunohistochemical staining of Reg IV (original magnification ×200).

IV expression more frequently than poorly differentiated adenocarcinoma (P b .005) (Table 1).

3.3. Relationship between Reg IV and Cdx2 expressions The goblet cells of intestinal metaplasia with Reg IV expression in the mucosa of adenomyomatosis and in the surrounding mucosa of the tumoral lesion showed Cdx2 expression. In GBC, 17 (28%) of 61 cases showed Cdx2 expression, in which positive Cdx2 staining was seen in the nuclei of carcinoma cells (Fig. 4). The expression of Reg IV significantly correlated with Cdx2 expression (P b .005): 15 (88%) of 17 cases with Cdx2 staining showed positive Reg IV expression, and 25 of 44 cases with negative Cdx2 expression showed negative Reg IV expression.

3.4. Prognostic impact of Reg IV expression in GBC Fig. 1 Comparison of Reg IV transcript levels in 10 normal gallbladder tissues and 31 gallbladder carcinomas by quantitative RT-PCR.

Fig. 5 shows the comparative survival of patients with advanced GBC after surgical resection, according to immunohistochemical Reg IV expression. Patients with

1690

H. Tamura et al. Table 1 Reg IV–positive ratio according to depth of invasion and histologic type of gallbladder carcinoma Reg IV Reg IV Total % of (−) (+) positive cases Depth of invasion a T1a T1b T2 ≥T3

0 2 8 17

Histologic type Tubular adenocarcinoma G1 and G2 a 12 G3 a 11 Adenosquamous carcinoma 4 Total 27

6 1 15 12

6 3 23 29

100 33 65 41

31 b 3 0 34

43 14 4 61

72 21 0 56

a

According to the pathologic tumor node metastasis classification [19]. Significantly increased Reg IV–positive rate compared with other histologic types of gallbladder carcinoma. b

Reg IV levels than a cut-off of 3.0 ng/mL, which was defined on the basis of Reg IV levels of 20 healthy volunteers (the mean level ± 3 SD in these healthy controls). In all 4 cases with elevated serum Reg IV levels, positive Reg IV

Fig. 3 Expression of Reg IV in invasive gallbladder adenocarcinoma: Reg IV is observed in goblet-like cell vesicles (A) or in the cytoplasm of cancer cells (B) (original magnification ×200).

early GBC were excluded, because early GBC is expected to be cured by cholecystectomy alone [24]. Survival in patients with positive Reg IV expression was significantly better than that in patients with negative Reg IV expression. Among clinicopathologic variables possibly affecting prognosis, univariate analysis showed that hepatic parenchymal invasion, bile duct invasion, and neural invasion were significant negative prognostic factors. Age and sex of patients, cancer differentiation, invasion to the stomach or intestines, vascular invasion, lymph node involvement, and presence of cancer cells in the resection margin did not influence survival. The 4 significant prognostic factors determined by univariate analysis were included in a subsequent multivariate analysis. This analysis revealed negative Reg IV expression, as well as hepatic parenchymal invasion, to be independently associated with a poor prognosis (Table 2).

3.5. Serum Reg IV expression in GBC Before surgical resection, 4 (33%) of 12 patients with GBC, including a patient with early GBC, had higher serum

Fig. 4 Expression of Reg IV and Cdx2 in a representative invasive gallbladder adenocarcinoma. A, Hematoxylin and eosin staining; B, Reg IV; C, Cdx2 (original magnification ×400).

Reg IV expression in the gallbladder and gallbladder carcinoma

Fig. 5 Comparative patient survival of advanced GBC after surgical resection, according to immunohistochemical Reg IV expression. Patients with higher Reg IVexpression had significantly better survival after surgical resection.

immunostaining was observed in GBC tissues, although in 4 other cases with positive Reg IV immunostaining, serum Reg IV concentration was not elevated. None of the 4 patients with GBC showing negative Reg IV immunostaining had elevated serum Reg IV levels. Elevated serum Reg IV levels in 3 of these 4 patients decreased to within the cut-off value at least 4 weeks after the resection of their tumors. In contrast, serum Reg IV concentration was not elevated in 12 patients with cholelithiasis or adenomyomatosis, although in 2 of these 12 cases, the intestinal metaplasia with positive Reg IV immunostaining was observed.

4. Discussion Gallbladder carcinoma has been considered to arise from epithelial dysplasia, and dysplasia appears arising from the metaplasia. Molecular analysis supported this metaplasiadysplasia-carcinoma sequence, in which loss of heterozygosity (LOH) in the p53 gene, 9p LOH, and LOH in the DCC gene could be detected even in the metaplastic/hyperplastic epithelium of the gallbladder, and the pattern of allele loss in invasive carcinoma and in corresponding metaplasia or dysplasia was identical [25]. Two types of metaplasia are observed in the gallbladder mucosa, usually associated with chronic cholecystitis: the antral metaplasia and intestinal metaplasia. Both types of metaplasia may be associated with gallbladder carcinogenesis [11,14]. The present study showed that Reg IV mRNA and protein were expressed in more than half of GBC cases, including early disease. Immunohistochemical analysis revealed that Reg IV expression was found in goblet-like cell vesicles or in the cytoplasm of cancer cells, indicating that Reg IV expression might be associated with GBC having an intestinal phenotype. The correlation of Reg IV with Cdx2 expression supported this hypothesis. Furthermore, although we observed intestinal metaplasia in a few cases without malignancy, Reg IV expression was found not only in the cancer cells but also in the intestinal metaplastic epithelium

1691

of patients with adenomyomatosis, whereas this was never observed in the antral metaplasic epithelium or normal epithelium in the gallbladder. These results suggest that Reg IV may be involved in the development of GBC through intestinal metaplasia. Although the roles of Reg IV in cancerous tissues have not been completely clarified, recent reports indicated that Reg IV may function as a growth factor and as an antiapoptotic factor in gastric, colon, and pancreatic cancers [4,21,26,27]. It is possible that Reg IV participates in carcinogenesis of GBC with an intestinal phenotype through control or influence in cell proliferation and antiapoptosis effect. In one group of GBC, however, Reg IV expression decreased according to the degree of histologic dedifferentiation. Frequency with lower Reg IV protein expression in poorly differentiated compared to well-differentiated GBC suggested that Reg IV expression may be involved in the progression of carcinogenesis from a well- to a lessdifferentiated phenotype. Similarly to Reg I, which is another Reg family protein, expressed in intrahepatic cholangiocarcinoma [28], the loss of Reg IV expression may be involved in more malignant characters. Indeed, survival after surgical resection among patients with Reg IV–positive GBC was better than among patients with Reg IV–negative GBC. Few studies regarding the relationship between Reg IV expression and survival have been reported for some cancers. Gu et al [3] showed the association of Reg IVexpression with androgen independence and metastasis in prostate cancer. These results suggested the negative prognostic utility of Reg IV expression, although the association of Reg IV expression with cancer recurrence or patient survival was not reported. Mitani et al [21] reported that no statistically significant prognostic effect of Reg IV was found in patients with gastric cancer. In the present study, negative Reg IV protein expression in GBC was independently associated with a poor prognosis. These results suggest that the contribution of Reg IV expression to the outcome of patients with malignancy may vary according to the type of cancer. The reason why in our series of patients those with higher Reg IV expression had favorable prognosis may, at least in part, be attributed to the association of Reg IV expression with an intestinal phenotype of this tumor. Previous reports showed that serum Reg IV levels were elevated in patients with pancreatic cancer [4] and gastric cancer [21], even in patients with early-stage cancer. These Table 2 Multivariate analysis of prognostic factors in patients with gallbladder carcinoma Prognostic factors

Relative risk (95% confidence interval)

P values

Hepatic parenchymal invasion Bile duct invasion Neural invasion Negative Reg IV expression

2.0 (1.0-4.0) 1.6 (0.7-3.5) 1.4 (0.6-3.2) 2.1 (1.1-4.0)

.04 .26 .38 .03

1692 reports have proposed that Reg IV is a good serum marker for detecting early stage of cancers. In the current study, we demonstrated that the elevated serum Reg IV levels could be detected in 33% of patients with GBC, whereas none of the patients with benign diseases of the gallbladder had elevated serum Reg IV levels. Taken together with other indicators, elevated serum Reg IV concentration may be a marker for detection of certain types of Reg IV–positive cancers, even in early stages, but specificity for determining the possible existence of a tumor is questionable. Rather, elevated serum Reg IV levels might be useful in distinguishing GBC from chronic cholecystitis and adenomyomatosis, which is sometimes difficult in x-ray images [29,30]. In addition, Reg IV concentration in sera may be a marker for GBC with an intestinal phenotype and for prediction of relatively favorable prognosis. The fact that all GBC cases and benign diseases of the gallbladder with positive Reg IV immnostaining did not necessarily show elevated serum Reg IV levels might be due to different sensitivity between both methods. Alternatively, the reason why only patients with GBC, but not those with benign diseases of the gallbladder even with positive Reg IV immunostaining, had elevated serum Reg IV levels may be that there are some mechanisms of a neoplastic amplification of Reg IV expression. These points require further study. In conclusion, we demonstrated that aberrant Reg IV mRNA and protein are expressed in GBC. Reg IV protein may be involved in GBC carcinogenesis through intestinal metaplasia and is associated with relatively favorable prognosis in patients with GBC after surgical resection. We were able to detect Reg IV protein in sera from patients with positive Reg IV immunostaining, suggesting the possibility that Reg IV is a novel serum biomarker for GBC.

References [1] Hartupee JC, Zhang H, Bonaldo MF, Soares MB, Dieckgraefe BK. Isolation and characterization of a cDNA encoding a novel member of the human regenerating protein family: Reg IV. Biochim Biophys Acta 2001;1518:287-93. [2] Oue N, Mitani Y, Aung PP, et al. Expression and localization of Reg IV in human neoplastic and non-neoplastic tissues: Reg IV expression is associated with intestinal and neuroendocrine differentiation in gastric adenocarcinoma. J Pathol 2005;207:185-98. [3] Gu Z, Rubin MA, Yang Y, et al. Reg IV: a promising marker of hormone refractory metastatic prostate cancer. Clin Cancer Res 2005;11:2237-43. [4] Takehara A, Eguchi H, Ohigashi H, et al. Novel tumor marker REG4 detected in serum of patients with resectable pancreatic cancer and feasibility for antibody therapy targeting REG4. Cancer Sci 2006;97: 1191-7. [5] Zhang Y, Lai M, Lv B, et al. Overexpression of Reg IV in colorectal adenoma. Cancer Letters 2003;200:69-76. [6] Miyazaki M, Ito H, Ambiru S, et al. Radical surgery for advanced gallbladder carcinoma. Br J Surg 1996;83:478-81. [7] Donohue JH. Present status of the diagnosis and treatment of gallbladder carcinoma. J Hepatobiliary Pancreat Surg 2001;8:530-4. [8] Misra S, Chaturvedi A, Misra NC, Sharma ID. Carcinoma of the gallbladder. Lancet Oncol 2003;4:167-76.

H. Tamura et al. [9] Roa I, De Aretxabala X, Araya JC, Roa J. Preneoplastic lesions in gallbladder cancer. J Surg Oncol 2006;93:615-23. [10] Lewis JT, Talwalkar JA, Rosen CB, Smyrk TC, Abraham SC. Prevalence and risk factors for gallbladder neoplasia in patients with primary sclerosing cholangitis: evidence for a metaplasia-dysplasiacarcinoma sequence. Am J Surg Pathol 2007;31:907-13. [11] Duarte I, Llanos O, Domke H, Hatz C, Valdivieso V. Metaplasia and precursor lesions of gallbladder carcinoma. Frequency, distribution, and probability of detection in routine histologic samples. Cancer 1993;72:1878-84. [12] Sakamoto H, Mutoh H, ido K, Satoh K, Hayakawa H, Sugano K. A close relationship between intestinal metaplasia and Cdx2 expression in human gallbladders with cholelithiasis. HUM PATHOL 2007;38: 66-71. [13] Fernandes JEV, Franco MIF, Suzuki RK, Tavares NM, Bromberg SH. Intestinal metaplasia in gallbladders: prevalence study. Sao Paulo Med J 2008;126:220-2. [14] Mukhopadhyay S, Landas SK. Putative precursors of gallbladder dysplasia: a review of 400 routinely resected specimens. Arch Pathol Lab Med 2005;129:386-90. [15] Yuasa Y. Control of gut differentiation and intestinal-type gastric carcinogenesis. Nat Rev Cancer 2003;3:592-600. [16] Wu XS, Akiyama Y, Igari T, et al. Expression of homeodomain protein CDX2 in gallbladder carcinomas. J Cancer Res Clin Oncol 2005;131: 271-8. [17] Silberg DG, Swain GP, Suh ER, Traber PG. Cdx1 and cdx2 expression during intestinal development. Gastroenterology 2000;119:961-71. [18] Nakata K, Nagai E, Ohuchida K, et al. REG4 is associated with carcinogenesis in the ‘intestinal’ pathway of intraductal papillary mucinous neoplasms. Mod Pathol 2009;22:460-8. [19] Sobin LH, Wittekind Ch, editors. International Union Against Cancer: TNM classification of malignant tumors. 6th ed. New York: WileyLiss; 2002. [20] Moskaluk CA, Zhang H, Powell SM, Cerilli LA, Hampton GM, Frierson HF. Cdx2 protein expression in normal and malignant human tissues: an immunohistochemical survey using tissue microarrays. Mod Pathol 2003;16:913-9. [21] Mitani Y, Oue N, Matsumura S, et al. Reg IV is a serum biomarker for gastric cancer patients and predicts response to 5-fluorouracil-based chemotherapy. Oncogene 2007;26:4383-93. [22] Matsumoto K, Mizoshita T, Tsukamoto T, et al. Cdx2 expression in pancreatic tumors: relationship with prognosis of invasive ductal carcinomas. Oncol Rep 2004;12:1239-43. [23] Adsay NV, Merati K, Basturk O, et al. Pathologically and biologically distinct types of epithelium in intraductal papillary mucinous neoplasms. Delineation of an “intestinal” pathway of carcinogenesis in the pancreas. Am J Surg Pathol 2004;28:839-48. [24] de Aretxabala X, Roa I, Burgos L. Gallbladder cancer, management of early tumors. Hepatogastroenterology 1999;46:1547-51. [25] Wistuba II, Sugio K, Hung J, et al. Allele-specific mutations involved in the pathogenesis of endemic gallbladder carcinoma in Chile. Cancer Res 1995;55:2511-5. [26] Bishnupuri KS, Luo Q, Murmu N, Houchen CW, Anant S, Dieckgraefe BK. Reg IV activates the epidermal growth factor receptor/Akt/AP-1 signaling pathway in colon carcinomas. Gastroenterology 2006;130:137-49. [27] Nanakin A, Fukui H, Fujii S, et al. Expression of the Reg IV gene in ulcerative colitis. Lab Invest 2007;87:304-14. [28] Harada K, Zen Y, Kanemori Y, et al. Human REG I gene is upregulated in intrahepatic cholangiocarcinoma and its precursor lesions. Hepatology 2001;33:1036-42. [29] Gore RM, Yaghmai V, Newmark GM, Berlin JW, Miller FH. Imaging benign and malignant disease of the gallbladder. Radiol Clin North Am 2002;40:1307-23. [30] Ching BH, Yeh BM, Westphalen AC, Joe BN, Qayyum A, Coakley FV. CT differentiation of adenomyomatosis and gallbladder cancer. AJR 2007;189:62-6.