Evaluation of REG4 for Early Diagnosis and Prognosis Of

Human Pathology (2011) 42, 1401–1409

www.elsevier.com/locate/humpath

Original contribution Evaluation of REG4 for early diagnosis and prognosis of gastric cancer☆ Hou-Quan Tao MD a,⁎, Xu-Jun He MB a, Ying-Yu Ma MS a, Hui-Ju Wang MS a, Ying-Jie Xia MB a, Zai-Yuan Ye MS a, Zhong-Sheng Zhao MS b aKey Laboratory of Gastroenterology of Zhejiang Province, Hangzhou, 310014, China bDepartment of Pathology, Zhejiang Provincial People's Hospital, Hangzhou, 310014, China

Received 2 April 2010; revised 30 July 2010; accepted 3 August 2010

Keywords: Summary We explored the correlation between the development of gastric cancer and the concentration REG4; of REG4 and hence the suitability of REG4 as an indicator of the prognosis of patients with GC. Real- Gastric cancer time polymerase chain reaction was conducted to detect REG4 messenger RNA expression. The amount of the REG4 protein was measured by immunohistochemistry staining of tissue and enzyme-linked immunosorbent assay of serum. Serum carcinoembryonic antigen and carbohydrate antigen 19-9 concentrations were measured using a commercial automated immunoassay. Real-time polymerase chain reaction results confirmed that REG4 was significantly up-regulated in gastric cancer compared with paired normal mucosa (P b .001). Immunohistochemistry staining revealed that high expression of REG4 correlated with diffuse type, poor differentiation, lymph node metastasis, distant metastasis, and TNM stage III or IV. The mean survival time for patients in the REG4-positive group was significantly less than that in the REG4-negative group (P = .013). The percentage of serum samples that were REG4 positive was 44.0%, which was higher than that for serum carcinoembryonic antigen (P = .039) or carbohydrate antigen 19-9 (P = .012) in TNM stage I and was significantly higher (P = .031) than that in TNM stage II. Thus, REG4 may be not only a prognostic indicator but also a better serum marker than carcinoembryonic antigen and carbohydrate antigen 19-9 for early diagnosis of gastric cancer. © 2011 Elsevier Inc. All rights reserved.

1. Introduction habits. GC in China constitutes approximately 33% of all worldwide GC cases [1]. As in all cancers, early detection Gastric cancer (GC) is one of the most common human remains the most promising approach to improve the long- cancers and is affected by environmental and behavioral term survival rate. Assessment of tumor markers in serum may be useful for this purpose. There are 2 available tumor markers for GC, carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9). However, neither is ☆ This work was carried out with the support of grants from Zhejiang an appropriate option for early screening, because preoper- Provincial Program for the Cultivation of High-Level Innovative Health ative positivity for these markers is highly reliant on tumor Talents and the Natural Science Foundation of Zhejiang Province (Grant No. Y207763). stage at the time of detection [2]. Establishment of screening ⁎ Corresponding author. strategies through the development of novel serum markers E-mail address: [email protected] (H. -Q. Tao). that are more specific and more sensitive to GC is thus

0046-8177/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2010.08.023 1402 H. -Q. Tao et al. urgently needed. Genes encoding transmembrane/secretory 2. Materials and methods proteins expressed specifically in cancers may be ideal diagnostic biomarkers [3]. The REG gene family belongs to 2.1. Blood Samples and isolation of serum the calcium-dependent lectin gene superfamily, which is mainly involved in hepatic, pancreatic, gastric, and intestinal Blood samples were obtained from 47 male and 23 female cell proliferation or differentiation. healthy donors (22-68 years old) and 65 male and 27 female The human REG gene family consists of 4 secretory and GC patients (32-86 years old) who were inpatients at the structurally unique proteins that are expressed in different Zhejiang Provincial People's Hospital between January and tissues and play different biological roles [4-6]. REG4,a June 2008. Samples were centrifuged for 10 minutes at member of the REG gene family, which was originally 1000 × g in a swing bucket rotor at 4°C, and the serum was identified by high-throughput sequencing of a complemen- stored in cryovials at −80°C until analysis. Tumor staging tary DNA (cDNA) library derived from a patient with was conducted according to the TNM system [19]. inflammatory bowel disease, maps to chromosome 1 and encodes a 158-amino-acid protein that includes a signal peptide of 22 amino acids and a conserved calcium- 2.2. Tissue samples dependent recognition domain [7]. Although various normal tissues express REG4, the concentrations in normal tissues One hundred ninety-two paraffin specimens of GC (138 are much lower than those in cancerous tissues [8]. In tumor male and 54 female patients aged 29-82 years) were acquired tissues of the gastrointestinal tract and pancreas or in cases in from Zhejiang Provincial People's Hospital, having been which these tissues are inflamed, REG4 can be seen to be collected from January 2000 to December 2001. All cases overexpressed in situ or in ectopias [9]. In colorectal were classified according to the World Health Organization carcinoma (CRC), up-regulation of REG4 was associated Pathological Classification of Tumors. There were 68 cases with tumor differentiation, stage, and lymph node metastasis of well or moderately differentiated tumors and 124 cases of [8,10]. High-throughput analysis showed that REG4 was poor or no differentiation. Histologically, there were 12 cases overexpressed in GC, CRC, pancreatic cancer, and prostate of papillary adenocarcinoma, 112 cases of tubular adeno- cancer but not in lung and breast cancer, suggesting that carcinoma, 58 cases of mucinous adenocarcinoma, and 10 REG4 may be tissue-specific and serve as a marker for cases of signet-ring cell carcinoma. There were 76 cases of tumors of the digestive system and prostate [11-14]. The intestinal histologic type and 116 cases of diffuse histologic REG4 protein is a growth promoter and may function as an type according to the Lauren classification. In 102 cases, antiapoptotic factor early in the development of GC [15,16]. there were lymph node metastases, whereas in 90 cases, no In addition, REG4 is overexpressed in prostate cancer, where such metastases were apparent. Eighteen cases were TNM it is associated with progression and metastasis of hormone- stage I, 45 cases were stage II, 62 cases were stage III, and 67 resistant disease [17]. cases were stage IV. None of the patients had received any Quantitative real-time reverse transcriptase-polymerase radiotherapy or chemotherapy in advance of the operation. chain reaction (PCR) has been used to investigate REG4 Specimens were fixed in formalin and embedded in paraffin. messenger RNA (mRNA) expression in GC cell lines and All patients were followed for more than 5 years. tissues. Results indicate that REG4 is highly overexpressed Thirty-seven fresh specimens from patients with GC were in GC cell lines and tissues established from peritoneal acquired from Zhejiang Provincial People's Hospital from − dissemination, and the quantity of REG4 correlated with wall January 2008 to December 2008 and stored at 80°C until penetration, suggesting that REG4 is involved in the use to detect relative REG4 mRNA expression by real-time peritoneal dissemination of GCs and may be a novel marker PCR. Surrounding normal gastric mucosa also was obtained for such disease [14,18]. and studied. Early detection remains the most promising approach to The project was approved by the ethics committee of improving the long-term survival of patients with GC, and Zhejiang Provincial People's Hospital. serum markers seem to play important roles in its diagnosis. REG4, a secretory protein, could be an alternative serum 2.3. REG4 mRNA expression analysis marker. Few researchers have studied the correlation between REG4 expression and clinicopathologic features Total RNA was isolated with Trizol (Invitrogen, Carls- of GC. In this study, we used immunohistochemical staining bad, CA). A total of 2 μg RNA was reverse transcribed using and enzyme-linked immunosorbent assays (ELISAs) to the SuperScript II RNase-Reverse Transcriptase System determine REG4 expression in sera and tissue and also (Invitrogen). The cDNA was then subjected to real-time PCR investigated the serum concentrations of CEA and CA19-9 with primers specific for REG4. A 0.4-μL aliquot of each using a commercially available automated immunoassay. cDNA and 0.4 μmol/L of forward and reverse primers were The correlation between REG4 expression and tumor added to SYBR Premix ExTaq (Perfect Real Time; TaKaRa invasion and metastasis and the significance of serum Bio Inc, Otsu, Japan) in a total volume of 25 μL. The REG4 concentrations were analyzed. amplification protocol was as follows: denaturation at 95°C REG4 in GC diagnosis and prognosis 1403 for 4 minutes and 40 cycles of 95°C for 10 seconds, 55°C for was added to a 96-well plate that had been coated with REG4 20 seconds, and 72°C for 20 seconds. The sequence of the antibody and incubated at 37°C for 2 hours. Then, 100 μLof forward primer for REG4 was 5′-GCC CTT AGA GTC TTG biotin-antibody working solution was added to each well, and GTT GC-3′ and that of the reverse primer was 5′-GCC ATC the plate was incubated at 37°C for 1 hour, followed by TTC CTC CTA CCC TG-3′. The sequence of the primers addition of 100 μL of horseradish peroxidase–avidin used for the control GAPDH was 5′- TGA AGG TCG GAG working solution. After incubation at 37°C for 1 hour, the TCA ACG G-3′ (forward) and 5′-CTG GAA GAT GGT contents of each well were aspirated and replaced with 90 μL GAT GGG ATT-3′ (reverse; Invitrogen). The amount of of the chromogenic peroxidase substrate tetramethylbenzi- REG4 mRNA relative to GAPDH was calculated as the dine. The plate was incubated for 30 minutes at 37°C in the −ΔCt average 2 , where ΔCt = Ct − CtGAPDH. dark, and the reaction was terminated by adding 50 μLof “stop solution” (provided by the manufacturer). The 2.4. Immunohistochemical staining absorbance at 450 nm was recorded using a microplate reader.

Briefly, each tissue section was dewaxed, rehydrated, and 2.6. Measurement of serum CEA and CA19-9 incubated with fresh 3% (vol/vol) H2O2 for 10 minutes. High-pressure antigen retrieval from the tissue was carried Serum CEA and CA19-9 were measured by using a out in 0.01 mol/l citrate buffer (pH 6.0). Sections were commercially available automated immunoassay method incubated with 10% (vol/vol) normal goat serum for (Modular Analytics; Roche Diagnostics, Basel, Switzerland) 15 minutes at room temperature. After rinsing with according to the manufacturer's instructions. The upper phosphate-buffered saline (PBS), slides were incubated limits of detection for this assay are 5.0 ng/mL for CEA and overnight at 4°C with rat antihuman REG4 monoclonal 37 U/mL for CA19-9. antibody (1:60 dilution in PBS; R&D Systems, Minneapolis, MN). After rinsing with PBS, tissue sections were incubated 2.7. Statistical analysis for 20 minutes at room temperature with biotin-labeled secondary antibody. After rinsing with PBS, tissue sections Statistical analysis of the data was performed using SPSS were incubated for 20 minutes at room temperature with version 13.0 (SPSS, Chicago, IL). The χ2 test, Wilcoxon – horseradish peroxidase linked goat antirat antibody W test, Mann-Whitney U test, Cox regression analysis, and (Zymed, San Francisco, CA). Subsequently, sections were Kaplan-Meier survival analysis were conducted as appropriate. stained with 3,3-diaminobenzidine and counterstained with A difference was considered significant when P b .05. hematoxylin, then dehydrated and mounted. The stained sections were evaluated by 2 independent investigators using a Nikon light microscope (Nikon Corporation, Tokyo, Japan), and representative photographs 3. Results were taken. The intensity of REG4 immunoreactivity was scored on a scale of 0 to 3+ (0 for no staining, 1+ for weak 3.1. Expression of REG4 mRNA immunoreactivity, 2+ for moderate immunoreactivity, and 3+ for strong immunoreactivity). The proportion of cells that To detect REG4 mRNA expression, a total of 37 paired demonstrated REG4 staining within the normal/cancerous fresh specimens of GC and surrounding normal mucosa were region of a section was scored as follows: 0 for less than 5% analyzed using real-time PCR. Relative to GAPDH, REG4 of cells positive, 1+ for 5% to 25%, 2+ for 26% to 50%, 3+ was significantly up-regulated in 75.7% of GCs (28/37) and for 51% to 75%, and 4+ for 76% to 100%. The staining down-regulated in the remainder (P b .001). Fig. 1A shows intensity and the percent immunoreactivity scores were then that the mean concentration of REG4 was significantly multiplied to obtain a composite score. The values of the higher in GC than in normal mucosa (P b .001; Wilcoxon composite score ranged from 0 to 12, with 0 to 3 defined as test). Fig. 1B displays the relative REG4 concentrations in negative and 4 or higher categorized as positive. the 37 paired specimens.

2.5. Determination of serum REG4 concentrations 3.2. Immunohistochemical staining for REG4 by sandwich ELISA A total of 192 GCs were examined for REG4 immunos- The concentration of REG4 in the serum of patients was taining. Yellow-brownREG4granuleswereobserved determined with the REG4 ELISA Kit (Cusabio Biotech Co mainly in the cytoplasm (Fig. 2B, D, F, and H). Expression Ltd, Newark, NJ). Sera from patients with GC were tested, was detected both in GC tissues (Fig. 2B, D, and F) and in with sera from healthy donors used as controls. A 100-μL lymph nodes with metastases (Fig. 2H). The proportion of volume of either standard (provided by manufacturer), REG4-positive specimens was 53.1% (102/192) in the GC sample (GC or healthy control), or sample diluent (blank) specimens. The proportion of REG4-positive samples 1404 H. -Q. Tao et al.

Fig. 1 Real-time PCR results of GC and normal tissues. A, Concentration of REG4 relative to GAPDH was higher in GC than in normal tissue (P b .05). B, Relative REG4 concentrations in 37 paired specimens. correlated statistically with the Lauren classification, differ- samples positive for REG4 was 69.6% (39/56) in specimens entiation, lymph node metastasis, distant metastases, and with distant metastasis, as opposed to those without distant TNM stage (Table 1). The frequency of REG4 positivity in metastasis (46.3%, 63/136; P b .001). Also, REG4 was patients with the diffuse type of GC (62.1%; 72/116) was detected in 33.3% of specimens (21/63) in early stages (TNM significantly higher than in those with the intestinal type I + II), much lower in than those at later stages (TNM III + (39.5%; 30/76; P = .003). The rate of samples positive for IV, 62.8%; 81/129; P b .001). REG4 was 62.9% (78/124) in poorly differentiated GCs, The mean survival time in groups positive for REG4 was higher than in those with well- to moderately differentiated 57.05 ± 6.01 months, which was significantly lower than in tumors (35.3%; 24/68; P b .001). The REG4 protein was those negative for REG4 (79.60 ± 6.26 months; P = .013). detected in 66.7% (68/102) of GC specimens with lymph The 5-year survival time for the group that was positive for node metastases, which is higher in than those without REG4 expression (46.0%) was significantly lower than that metastases (37.8%; 34/90; P b .001). The percentage of in the negative group (68.9%; P = .013; Fig. 3). Cox

Fig. 2 Immunohistochemical analysis of REG4 in GC. A, C, E, G, I, Original magnification ×40. B, D, F, H, J, Original magnification ×400. A, Immunostaining of REG4 (yellow-brown granules, mainly in the cytoplasm) in poorly differentiated adenocarcinoma. C, Immunostaining of REG4 in moderately differentiated adenocarcinoma. E, Immunostaining of REG4 in gastric adenocarcinoma with muscle layer invasion. G, Immunostaining of REG4 in lymph node with metastasis in gastric adenocarcinoma. I, Negative control for immunostaining of REG4, with PBS replacing primary antibody against REG4. REG4 in GC diagnosis and prognosis 1405 1406 H. -Q. Tao et al.

Table 1 Correlation between REG4 expression and clinicopathologic features of GC Groups Cases Cases positive Cases negative Frequency rate of P for REG4 for REG4 positive REG4 Sex .873 Male 138 74 64 53.6% Female 54 28 26 51.9% Tumor diameter (cm) .885 b4 80 42 38 52.5% ≥4 112 60 52 53.6% Lauren classification .003 Diffuse type 116 72 44 62.1% Intestinal type 76 30 46 39.5% Differentiation .000 Well or moderate 68 24 44 35.3% Poor or no 124 78 46 62.9% Histology classification .463 Papillary adenocarcinoma 12 6 6 50.0% Tubular adenocarcinoma 112 64 48 57.1% Mucinous adenocarcinoma 58 26 32 44.8% Signet-ring cell carcinoma 10 6 4 60.0% Lymph node metastasis .000 Yes 102 68 34 66.7% No 90 34 56 37.8% Distant metastasis .000 Yes 56 39 17 69.6% No 136 63 73 46.3% TNM stage .000 I + II 63 21 42 33.3% III + IV 129 81 48 62.8%

regression analysis showed that Lauren classification and 3.3. Quantitative ELISA for determination of REG4 TNM stage were independent prognostic factors; however, concentrations in serum REG4 expression was not independent (Table 2). The serum concentrations of REG4 in persons with and without GC were analyzed by the Mann-Whitney U test, and the REG4 concentration was found to be higher in GC patients (2.125 ± 0.184 ng/mL) than in healthy donors (0.767 ± 0.061 ng/mL; P b .01). Receiver operating characteristic (ROC) curve analysis revealed that the area under the curve was 0.798, which indicates that the test was accurate in classifying cases as “cancerous” or “normal” (Fig. 4).

3.4. Diagnostic sensitivity of serum REG4, CEA, and CA199 with respect to TNM stage

The CEA and CA19-9 concentrations were measured in the same serum samples. The serum concentrations of REG4, CEA, and CA199 from all samples measured are supplied as Supplementary Table 1. To determine the sensitivity of elevated REG4 as a diagnostic test, we measured the serum REG4 concentrations of 70 healthy volunteers, and a cutoff value of 2.053 ng/mL for normal was defined, which was the Fig. 3 Kaplan-Meier survival curve of patients positive and mean value plus 2 standard deviations. Serum CEA negative for REG4 expression. concentrations greater than 5 ng/mL or serum CA19-9 REG4 in GC diagnosis and prognosis 1407

Table 2 Multivariate analysis as determined by Cox regression analysis Clinicopathologic parameters 95% confidential interval Hazard ratio P Lower Upper Lauren classification 1.244 6.854 2.920 .014 Histology classification 0.434 1.085 0.686 .107 TNM stage 1.422 4.845 2.625 .002 REG4 expression 0.925 4.103 1.948 .079 concentrations greater than 37 U/mL were considered in diagnosis of gastrointestinal signet-ring cell carcinoma [22]. positive for disease. The proportion of serum samples Quantitative reverse transcriptase PCR analysis revealed that positive for REG4 (44.0%) was significantly higher than approximately 50% of GCs overexpress REG4 [13].Our those for the expression of CEA (16.0%; P = .039) and study found that REG4 mRNA expression was up-regulated CA19-9 (8.0%; P = .012) in TNM stage I. Also, it was in carcinomas compared with surrounding normal gastric significantly higher than that of CEA (53.8% versus 7.7%; mucosa. Further investigation of REG4 protein expression in P = .031) in TNM stage II (Table 3), implying that REG4 192 GC specimens with immunohistochemical staining may be a better serum marker for early diagnosis of GC than revealed that the percentage of REG4-positive samples was are CEA and CA19-9. 53.1% (102/192) in the GC specimens, similar to the results reported by Oue et al [8]. The RELP protein is constitutively expressed in epithelial neuroendocrine cells of the small intestine and in parietal cells of the gastric mucosa. Up- 4. Discussion regulated expression of REG4 was demonstrated in the epithelial cells of inflammatory mucosa in ulcerative colitis The REG4 protein was overexpressed in most colorectal and Crohn disease, as well as in the regenerating epithelial adenocarcinomas, and its up-regulation was associated with borders of gastric ulcers and the metaplastic epithelium in the the formation of adenoma and adenocarcinoma [10,20,21]. antrum and the esophagus. These findings indicate that REG4 Recent research suggests that REG4 is expressed mainly in the usually plays an important role in gastric carcinogenesis. distal gastrointestinal tract, including the stomach, intestine, Zheng et al [23] also reported that REG4 expression colon, and pancreas. Staining of REG4 and claudin-18 can aid experienced up-regulation in gastric intestinal metaplasia

Fig. 4 A, ROC curve analysis of diagnostic sensitivity and specificity of the serum REG4 ELISA assay. The performance of the assay in discriminating patients with GC from those without was evaluated. The area under the curve was 0.798, suggesting that the test was accurate in distinguishing between the 2 groups. B, A graph of ELISA data showed serum REG4 concentrations were higher in GC patients than in normal persons (**P b .01). 1408 H. -Q. Tao et al.

Table 3 Diagnostic sensitivity of serum REG4, CEA, and CA199 with respect to TNM stage TNM stage Cases REG4-positive CEA-positive CA199-positive P value P value ratio ratio ratio (REG4 vs CEA) (REG4 vs CA199) I 25 44.0% (11/25) 16.0% (4/25) 8.0% (2/25) .039 .012 II 13 53.8% (7/13) 7.7% (1/13) 15.4% (2/13) .031 .125 III 40 45.0% (18/40) 27.5% (11/40) 30.0% (12/40) .210 .238 IV 14 57.1% (8/14) 42.9% (6/14) 21.4% (3/14) .688 .063 Using an established cutoff for healthy individuals (2.053 ng/mL), ≥2.053 ng/mL was positive for REG4, ≥5 ng/mL was positive for CEA, and ≥37U/mL was positive for CA199.

and adenoma and then down-regulation with malignant cinoma [27]. Another study reported that REG4 may be a transformation of gastric epithelial cells, which suggested serum biomarker for GC [28]. In our study, the mean that REG4 expression should be considered a good biomarker concentrations of serum REG4 were higher in GC patients for precancerous gastric lesions. In our study, the proportion than in healthy individuals (P b .01). The ROC curve of REG4-positive specimens among 192 GCs correlated analysis implied that the test was fairly accurate in with the Lauren classification, extent of histologic differen- classifying cases as “cancerous” or “normal,” suggesting tiation, lymph node metastasis, distant metastasis, and TNM that serum REG 4 may be an adequate serum marker for GC stage. The proportion of REG4-positive tumors was higher in as well. diffuse and poorly differentiated histologic types than in CEA has been widely accepted as a marker useful in the intestinal and well-differentiated types. Because we did not diagnosis and management of CRC. Kim et al [29] used have adequate samples of gastritis and benign tumors, we can radioimmunoassays to measure preoperative serum CEA only postulate that REG4 participates in the development of concentrations and found that GC patients with preoperative GC. The specific role of REG4 in gastric carcinogenesis needs serum CEA concentrations greater than 10.0 ng/mL had to be studied further. more prominent serosal invasion, much more lymph node Expression of REG4 mRNA has potential as a novel involvement, more advanced-stage tumors, and more poorly marker for detecting peritoneal dissemination in GC [18]. differentiated tumors than did the patients with preoperative Also, REG4 was reported to accelerate peritoneal metastasis serum CEA concentrations less than 5.0 ng/mL. The survival in GC and might therefore be a good marker to apply to lavage rate of GC patients with serum CEA concentrations greater fluids [24]. Our research also found that the number of REG4- than 10.0 ng/mL was lower than for those patients with positive samples was higher among specimens from patients serum CEA concentrations less than 5.0 ng/mL. with lymph node and distant metastasis than that in specimens Serum CA19-9 concentrations in patients with GC have from patients without evident disease dissemination. We been reported to be significant. The proportion of patients postulate that REG4 participates in the process of GC positive for serum CA19-9 after radical surgery was 4%, metastasis, but the mechanism by which this occurs needs to but for those whose disease could not be resected, it was be further studied. The proportion of REG4-positive samples 64.9% [30]. in TNM stages I + II was lower than in stages III + IV. Even ignoring the reliability of CEA and CA19-9 as Besides, REG4 may be involved in invasion and deterioration markers for detection of GC, CEA and CA19-9 are still of GC according to what we have discovered. As recent inappropriate for the detection of early-stage GC, because studies have discovered that REG4 expression is an preoperative positivity for these markers is highly dependent independent prognostic indicator of relapse after radical on the tumor stage at the time of detection [2]. Thus, it is very prostatectomy [25]; it might accelerate cell growth and important to search for novel markers that provide excellent correlated with the prognosis of adenoid cystic carcinoma of performance in the diagnosis of GC in advance of treatment. the salivary glands [26]. We also found that the mean survival Dysregulation of REG occurs early in tumorigenesis, and time and 5-year survival rate in groups positive for REG4 increased expression, especially of REG4, contributes to were both significantly lower than for those in the negative adenoma formation and leads to greater resistance to group. Therefore, our results suggest that the expression of apoptotic cell death in CRC [31]. We found that serum REG4, although it was not an independent prognostic marker REG4 concentrations were higher in GC patients than in of GC by multivariate Cox regression analysis, was healthy donors. Zhang et al [10] said that overexpression of significantly associated with a poor prognosis. REG4 probably is an early event in colorectal carcinogen- Serum markers seem to play more important roles in the esis, and detection of REG4 overexpression could be useful early detection of GC. REG4 is a promising marker for in the early diagnosis of carcinomatous transformation of screening for early-stage pancreatic ductal adenocarcinoma, adenoma. The number of specimens positive for serum and the neutralization of REG4 by an antibody may offer a REG4 expression was significantly higher than those for novel tool for the treatment of pancreatic ductal adenocar- serum CEA or CA19-9 in TNM stage I and was also REG4 in GC diagnosis and prognosis 1409 significantly higher than that for serum CEA in TNM stage matrix metalloproteinase-10 are novel prognostic factors in patients II, indicating that REG4 is a better marker for early detection with gastric cancer. Oncogene 2006;25:2546-57. [12] Lee JY, Eom EM, Kim DS, Ha-Lee YM, Lee DH. Analysis of gene of GC than are CEA and CA19-9. expression profiles of gastric normal and cancer tissues by sage. In summary, REG4 protein expression is representative of Genomics 2003;82:78-85. features of GC with metastases and poor differentiation and [13] Oue N, Hamai Y, Mitani Y, et al. Gene expression profile of gastric could be used as an indicator of prognosis of GC. Also, carcinoma: Identification of genes and tags potentially involved in REG4 is likely to be a better serum marker for the early invasion, metastasis, and carcinogenesis by serial analysis of gene expression. Cancer Res 2004;64:2397-405. diagnosis of GC than are CEA and CA19-9, although [14] Yasui W, Oue N, Ito R, Kuraoka K, Nakayama H. Search for new confirmation of its clinical utility requires a larger and more biomarkers of gastric cancer through serial analysis of gene expression comprehensive study. and its clinical implications. Cancer Sci 2004;95:385-92. [15] Sekikawa A, Fukui H, Fujii S, et al. Reg ialpha protein may function as a trophic and/or anti-apoptotic factor in the development of gastric cancer. Gastroenterology 2005;128:642-53. Supplementary data [16] Unno M, Itoh T, Watanabe T, et al. Islet beta-cell regeneration and reg genes. Adv Exp Med Biol 1992;321:61-6 [discussion 67-9]. Supplementary materials related to this article can be [17] Violette S, Festor E, Pandrea-Vasile I, et al. Reg IV, a new member of the regenerating gene family, is overexpressed in colorectal carcino- found online at doi:10.1016/j.humpath.2010.08.023. mas. Int J Cancer 2003;103:185-93. [18] Miyagawa K, Sakakura C, Nakashima S, et al. Overexpression of reg IV in peritoneal dissemination of gastric cancer and its potential as a novel marker for the detection of peritoneal micrometastasis. References Anticancer Res 2008;28:1169-79. [19] Wittekind C, Compton CC, Greene FL, Sobin LH. TNM residual [1] Pavlick AC, Gerdes H, Portlock CS. Endoscopic ultrasound in the tumor classification revisited. Cancer 2002;94:2511-6. evaluation of gastric small lymphocytic mucosa-associated lymphoid [20] Zhang Y, Lai M, Gu X, Luo M, Shao L. Reg IV, a differentially tumors. J Clin Oncol 1997;15:1761-6. expressed gene in colorectal adenoma. Chin Med J (Engl) 2003;116: [2] Kochi M, Fujii M, Kanamori N, et al. Evaluation of serum CEA and 918-22. CA19-9 levels as prognostic factors in patients with gastric cancer. [21] Zhang YW, Ding LS, Lai MD. Reg gene family and human diseases. Gastric Cancer 2000;3:177-86. World J Gastroenterol 2003;9:2635-41. [3] Buckhaults P, Rago C, St Croix B, et al. Secreted and cell surface [22] Sentani K, Oue N, Tashiro T, et al. Immunohistochemical staining of genes expressed in benign and malignant colorectal tumors. Cancer reg IV and claudin-18 is useful in the diagnosis of gastrointestinal Res 2001;61:6996-7001. signet ring cell carcinoma. Am J Surg Pathol 2008;32:1182-9. [4] Abe M, Nata K, Akiyama T, et al. Identification of a novel reg [23] Zheng HC, Xu XY, Yu M, Takahashi H, Masuda S, Takano Y. The family gene, reg iiidelta, and mapping of all three types of reg role of Reg IV gene and its encoding product in gastric carcinogenesis. family gene in a 75 kilobase mouse genomic region. Gene 2000; HUM PATHOL 2010;41:59-69. 246:111-22. [24] Kuniyasu H, Oue N, Sasahira T, et al. Reg IV enhances peritoneal [5] Broekaert D, Eyckerman S, Lavens D, et al. Comparison of leptin- and metastasis in gastric carcinomas. Cell Prolif 2009;42:110-21. interleukin-6-regulated expression of the rpap gene family: evidence [25] Ohara S, Oue N, Matsubara A, et al. Reg IV is an independent for differential co-regulatory signals. Eur Cytokine Netw 2002;13: prognostic factor for relapse in patients with clinically localized 78-85. prostate cancer. Cancer Sci 2008;99:1570-7. [6] Okamoto H. The reg gene family and reg proteins: with special [26] Sasahira T, Oue N, Kirita T, et al. Reg IV expression is associated with attention to the regeneration of pancreatic beta-cells. J Hepatobiliary cell growth and prognosis of adenoid cystic carcinoma in the salivary Pancreat Surg 1999;6:254-62. gland. Histopathology 2008;53:667-75. [7] Hartupee JC, Zhang H, Bonaldo MF, Soares MB, Dieckgraefe BK. [27] Takehara A, Eguchi H, Ohigashi H, et al. Novel tumor marker reg4 Isolation and characterization of a cDNA encoding a novel member of detected in serum of patients with resectable pancreatic cancer and the human regenerating protein family: Reg IV. Biochim Biophys Acta feasibility for antibody therapy targeting reg4. Cancer Sci 2006;97: 2001;1518:287-93. 1191-7. [8] Oue N, Mitani Y, Aung PP, et al. Expression and localization of REG [28] Mitani Y, Oue N, Matsumura S, et al. Reg IV is a serum biomarker for IV in human neoplastic and non-neoplastic tissues: Reg IV expression gastric cancer patients and predicts response to 5-fluorouracil-based is associated with intestinal and neuroendocrine differentiation in chemotherapy. Oncogene 2007;26:4383-93. gastric adenocarcinoma. J Pathol 2005;207:185-98. [29] Kim DY, Kim HR, Shim JH, Park CS, Kim SK, Kim YJ. Significance [9] Kamarainen M, Heiskala K, Knuutila S, Heiskala M, Winqvist O, of serum and tissue carcinoembryonic antigen for the prognosis of Andersson LC. Relp, a novel human reg-like protein with up-regulated gastric carcinoma patients. J Surg Oncol 2000;74:185-92. expression in inflammatory and metaplastic gastrointestinal mucosa. [30] Marrelli D, Roviello F, De Stefano A, et al. Prognostic significance of Am J Pathol 2003;163:11-20. CEA, CA 19-9 and CA 72-4 preoperative serum levels in gastric [10] Zhang Y, Lai M, Lv B, et al. Overexpression of reg IV in colorectal carcinoma. Oncology 1999;57:55-62. adenoma. Cancer Lett 2003;200:69-76. [31] Bishnupuri KS, Luo Q, Korzenik JR, et al. Dysregulation of reg gene [11] Aung PP, Oue N, Mitani Y, et al. Systematic search for gastric cancer- expression occurs early in gastrointestinal tumorigenesis and regulates specific genes based on sage data: Melanoma inhibitory activity and anti-apoptotic genes. Cancer Biol Ther 2006;5:1714-20.