Serum Macrophage Migration- Inhibitory Factor as a Diagnostic and Prognostic Biomarker for Gastric Cancer

Transcription

1 Serum Macrophage Migration- Inhibitory Factor as a Diagnostic and Prognostic Biomarker for Gastric Cancer Harry Hua-Xiang Xia, PhD 1, ; Yi Yang, PhD 1, ; Kent-Man Chu, MBBS, MS 2 ; Qing Gu, PhD 1 ; Yuan-Yuan Zhang, PhD 3 ; Hua He, MPhil 1 ; Wai Man Wong, MD, PhD 1 ; Suet-Yi Leung, MD 4 ; Siu-Tsan Yuen, MBBS 4 ; Man-Fung Yuen, MD, PhD 1 ; Annie O.O. Chan, MD, PhD 1 ; and Benjamin C.Y. Wong, MD, PhD 1 BACKGROUND: This study aimed to determine the potential diagnostic value of migration-inhibitory factor (MIF) for gastric cancer in patients presenting with dyspepsia and its prognostic value for gastric cancer. METHODS: A cohort of 97 patients with histologically confirmed gastric adenocarcinoma and 222 patients with dyspepsia were recruited. Enzyme-linked immunosorbent assay was used to measure serum MIF and carcinoembryonic antigen (CEA). RESULTS: The serum MIF concentrations were pg/ml and pg/ml, respectively, in gastric cancer patients and dyspeptic patients (P <.001). Serum MIF levels increased with the advancing gastric pathologies (P <.001). With the cutoff value of 3230 pg/ml, serum MIF had sensitivity, specificity, and accuracy of 83.5%, 92.3%, and 89.7%, respectively, in diagnosing gastric cancer, whereas the rates were 60.8%, 83.3%, and 76.5%, respectively, for serum CEA. Gastric cancer patients with serum MIF levels above 6600 pg/ml had a lower 5-year survival rate than those with serum MIF level below that level (P ¼.012). Higher serum CEA levels were also associated with poor survival. The prediction for 5-year survival was even better (P ¼.0001), using a combination of serum MIF and CEA. CONCLUSIONS: Serum MIF level, which correlates with gastric MIF expression, is a better molecular marker than CEA in diagnosing gastric cancer in patients presenting with dyspepsia. A combination of serum MIF and CEA predicts 5-year survival better than the individual test. Cancer 2009;115: VC 2009 American Cancer Society. KEY WORDS: macrophage migration-inhibitory factor, carcinoembryonic antigen, gastric cancer, diagnosis, prognosis. Corresponding author: Benjamin C.Y. Wong, MD, PhD, Department of Medicine, Queen Mary Hospital, University of Hong Kong, Hong Kong, China; Fax: (011) ; bcywong@hku.hk 1 Department of Medicine, University of Hong Kong, Hong Kong, China; 2 Department of Surgery, University of Hong Kong, Hong Kong, China; 3 Department of Microbiology and Microbial Engineering, School of Life Sciences, Fudan University, Shanghai, China; 4 Department of Pathology, University of Hong Kong, Hong Kong, China The authors wish to thank Mr. Ying Jie Liang, of Department of Pathology, Sun Yat-Sen University of Medical Sciences, Guangzhou, China for performing part of the experiments, Miss Teresa Tong, Miss Vicky Ho, and Miss Fiona Fung for data management, and Mr. Hess KL Cheung for technical support. Part of the results in the study has been presented at the Digestive Disease Week of the American Gastroenterological Association, Orlando, May 17-22, Gastroenterology 2003; 124: A556. y HHX Xia s current address is Senior Medical Scientific Expert of Novartis Pharmaceuticals Corporation, East Hanover, New Jersey; Y. Yang, postdoctoral fellow at McKusick-Nathans Institute of Genetic Medicine, Johns Hopkins University, Baltimore, Maryland; and H. He, English editor of the Chinese Journal of Cancer, Guangzhou, China. Received: December 10, 2008; Revised: March 4, 2009; Accepted: March 31, 2009 Published online August 14, 2009 in Wiley InterScience ( DOI: /cncr.24609, Cancer December 1,

2 Gastric cancer is the second most common fatal malignancy in the world, leading to more than 750,000 deaths annually. 1 Most gastric cancers are diagnosed at an advanced stage and the overall survival is very poor, usually less than 15% at 5 years in many countries, 2 although the 5-year survival has been significantly improved in some Asian countries to up to 50% because of the early detection programs. 3-4 Gastric cancer is associated with socioeconomic status, genetic, and environmental factors. 5 Macrophage migration-inhibitory factor (MIF) plays a pivotal role in inflammatory and immune-mediated diseases. 6 Moreover, MIF expression is increased in many cancers including melanoma, neuroblastoma, and cancers of prostate, liver, and breast, suggesting its involvement in carcinogenesis MIF expression in gastric and intestinal epithelial cells and its role in gastrointestinal diseases have been investigated In an animal study, we observed that MIF messenger ribonucleic acid (mrna) and protein were markedly up-regulated, and that blockade of MIF with the neutralizing antibody inhibited the up-regulation of MIF. 19 In addition, we demonstrated that H. pylori infection was associated with increased MIF expression by both gastric epithelial cells and inflammatory cells. 20 These findings indicate that MIF is a key inflammatory mediator and plays a pathogenic role in gastric inflammation. Moreover, we observed that both serum and epithelial MIF expression is progressively increased in H. pylori-induced gastritis, intestinal metaplasia, and gastric cancer, suggesting that MIF is involved in gastric carcinogenesis and may be a valuable biomarker for the early detection of gastric cancer and for the prediction of prognosis of gastric cancer patients. 22 Therefore, the aim of this study was to determine the potential clinical value of serum MIF in diagnosing gastric cancer in patients with dyspepsia and its prognostic value for gastric cancer patients. MATERIALS AND METHODS Patients A cohort of 97 patients (65 14 years, male/female, 61 of 36) from our previous study 23,24 were recruited into the present study. These patients were admitted between January 1, 1997, to September 30, 1998, to the Departments of Medicine and Surgery, Queen Mary Hospital with histologically confirmed gastric adenocarcinoma and Table 1. Demographic, Clinical, and Histological Findings of the Patients With Gastric Cancer (n¼97) Findings Age, y; meanstandard deviation No. (Range) 6514 Sex Men 61 Women 36 Survival, y, median (range) 1.28 ( ) Tumor size, cm; median (range) 4.5 (0.5-18) Tumor site Antrum/Incisura 67 Other sites 30 Differentiation Poor 1 Moderate 62 Well 34 Depth of invasion T1 6 T2 23 T3 48 T4 20 Lymph node metastasis N0 14 N1 30 N2 53 Peritoneal/Hepatic metastasis Yes 39 No 58 Stage I 12 II 16 III 23 IV 46 Operability Palliative/conservative 51 Operable 46 were followed up as previously reported. 23,24 The last date of follow-up for the survivors was June 30, 2003 (Table 1). The tumors were staged and classified according to the Japanese Research Society for Gastric Cancer, the World Health Organization, and Lauren s classification. 25,26 Serum samples were collected after histological confirmation of gastric cancer but before operation or initiation of chemotherapy. Venous blood samples were collected in plain tubes, allowed to clot, and within 1 hour of collection, centrifuged at 800 g for 10 minutes at 4 C to obtain the serum. The serum was removed, aliquoted, and stored at 70 C. 23,24 As controls, 222 patients (46 15 years, male/ female, 84 of 138) referred for upper endoscopy at the 5442 Cancer December 1, 2009

3 MIF as a Biomarker for Gastric Cancer/Xia et al Department of Medicine, Queen Mary Hospital between January 2002 and December 2002 because of dyspeptic symptoms and without endoscopic evidence of gastric cancers were recruited. None of the patients suffered from any other tumors or other obvious inflammatory diseases such as glomerulonephritis and rheumatoid arthritis, which are known to be associated with increased MIF expression. Serum samples were collected by using the technique, as described above, before the endoscopy, and they were stored at 70 C until use in this study. At endoscopy, 4 biopsies were obtained from each patient: 3 from the gastric antrum, and 1 from the body. One antral biopsy was used for an in-house rapid urease test. The other biopsies were fixed in formalin and embedded in paraffin for histological examinations. All patients gave informed written consents, and this project was approved by the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong Western Cluster, Hong Kong. Detection of H. pylori Infection and Histological Examinations The in-house rapid urease test has been shown to have a sensitivity of 99% and specificity of 100% for the detection of H. pylori. 27 Sections (4 lm) of the paraffin-embedded tissue were stained with hematoxylin and eosin, and Giemsa stain if necessary. Patients were defined to be H. pylori positive if positive by both rapid urease test and histology or by both histology at gastric antrum and body. Patients were defined to be H. pylori negative if all tests were negative. Patients positive by the rapid urease test, or histology at a single site only, were excluded from this study. For patients with gastric cancer, either histology or a locally validated serological test 28 was used for the detection of H. pylori infection. The updated Sydney system was adapted to assess the presence of chronic gastritis and intestinal metaplasia. 29 Enzyme-Linked Immunosorbent Assay for Detection of Serum MIF and Carcinoembryonic Antigen Serum MIF was measured in 2003 by an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer s instruction (R&D Systems, Minn). 22 The sensitivity of this assay was pg/ml of MIF. At the same time, serum carcinoembryonic antigen (CEA) was detected, using a commercial ELISA kit, according to the manufacturer s instructions (CanAg CEA EIA, CanAg Diagnostics AB, Gothenburg, Sweden), and the cutoff value of 5 ng/ml was recommended by the manufacturer, as well as in our previous studies, 23,24 and adopted in the present study. Statistic Analysis Data were expressed as the mean standard error of the mean (SEM). Statistical analysis was performed by using SPSS software (version 10.0, SPSS, Chicago, Ill). The independent samples t test, or the Mann-Whitney U test, was used to determine the differences in serum MIF concentration and gastric epithelial MIF expression among different groups. The Bivariate Correlation and Linear Regression assessed correlations between serum MIF concentrations and serum CEA concentrations. Receiver operating characteristic (ROC) curve analysis determined the cutoff serum MIF values for the diagnosis of gastric cancer. Cumulative survival rates were calculated by the Kaplan-Meier method. The associations of serum MIF with histological classifications, staging, and survivals were determined by the chi-square test, with Yates correction if required; odds ratios (OR) and 95% confidence intervals (CI) were calculated. Multivariate analysis by Cox regression in a forward stepwise manner was used to determine independent prognostic factors for survival. Two-tailed P values were calculated. The alpha level of significance was set at P <.05. RESULTS Endoscopic and Histological Findings of the Patients With Dyspepsia Among the 222 patients, 35 (15.8%) were diagnosed endoscopically as having peptic ulcer disease (5 gastric ulcer, 28 duodenal ulcer, and 2 both gastric and duodenal ulcer) and 187 as having nonulcer dyspepsia. Histologically, 112 patients had a normal gastric mucosa, 88 chronic gastritis, and 22 intestinal metaplasia. Overall, 101 (45.5%) patients were positive and 121 (54.5%) negative for H. pylori infection; none of patients were positive by a single test at a single site. Cancer December 1,

4 Diagnostic Value of Serum MIF Levels for Gastric Cancer The serum MIF concentrations were pg/ ml and pg/ml, respectively, in dyspeptic patients and those with gastric cancer (P <.001). In dyspeptic patients, the concentrations were pg/ml and pg/ml, respectively, in H. pylori positive and negative patients (P <.001), and and pg/ml, respectively, in patients with and without peptic ulcer disease (P <.001). The concentration increased with the progression of histological changes, ie, pg/ml, pg/ml, and pg/ml, respectively, in patients with normal gastric mucosa, chronic gastritis, and intestinal metaplasia (P <.001). There was a significant correlation between serum MIF levels and ages of patients (R ¼ 0.395, P <.001). However, this correlation disappeared after adjustment for histological diagnosis. Overall, serum MIF levels were significantly greater in male patients than in female patients ( pg/ ml vs pg/ml, P ¼.003). However, this difference disappeared after adjustment for the histological diagnosis. Serum CEA concentrations were and , respectively, in dyspeptic patients and those with gastric cancer (P <.001). There was a significant correlation between serum MIF and CEA concentrations (R ¼ 0.246, P <.001). When the entire cohort of patients and controls were assessed, the area under the ROC curve for serum MIF in diagnosing gastric cancer was (95% CI, ), which was greater than that for serum CEA (0.787 [95% CI, ]). According to the ROC curve, the serum MIF concentration of 3230 pg/ml was defined to be the optimal cutoff value for differentiating patients with gastric cancer and those with dyspepsia. With this cutoff value, the sensitivity, specificity, positive and negative predictive values, and accuracy of 83.5%, 92.3%, 82.7%, 92.8%, and 89.7%, respectively, were achieved (Table 2). The widely accepted cutoff value of 5 ng/ml of serum CEA was used for analysis in this study because the cutoff value (4.78 ng/ml) calculated according to the ROC curve was very close to this value. With this cutoff value, the sensitivity, specificity, positive and negative predictive values, and accuracy were 60.8%, Table 2. Values of Serum MIF (cutoff 3230 pg/ml) and CEA (cutoff 5 ng/ml) in Diagnosing Gastric Cancer (n¼319) MIF CEA MIF or CEA* MIF and CEA Sensitivity, % Specificity, % Positive predictive values, % Negative predictive values, % Positive likelihood ratio Negative likelihood ratio Accuracy, % MIF indicates macrophage migration-inhibitory factor; CEA, carcinoembryonic antigen. * Positive by either MIF or CEA serology. y Positive by both MIF and CEA serology. 83.3%, 61.5%, 83.0%, and 76.5%, respectively, which were inferior to those obtained by serum MIF (Table 2). Both sensitivity and negative predictive value improved when patients with either serum MIF or serum CEA positivity were considered to have a positive test, but the specificity was substantially decreased (Table 2). Prognostic Value of Serum MIF Levels for Gastric Cancer Patients All 97 patients with gastric cancer were further analyzed to determine the association of serum MIF and CEA concentrations with histological classifications, staging and survival. Of these patients, 15 (15.5%) survived for up to 5 years. Serum MIF concentrations were and (P ¼.022), respectively, in those who survived for 5 years and those who did not. According to the ROC curve, a concentration of 6600 pg/ml was calculated as an optimal cutoff value for predicting 5-year survival; patients with serum MIF higher than this cutoff were more likely to survive less than 5 years than those who had lower serum MIF concentrations (88% vs 57.1%; OR ¼ 5.63; 95% CI, ; P ¼.002). Moreover, serum MIF concentrations higher than this cutoff was significantly associated with stages III and IV (Table 3). Similarly, 79.7% of patients with serum CEA higher than 5 ng/ml and 50% of those with serum CEA lower than 5 ng/ml survived less than 5444 Cancer December 1, 2009

5 MIF as a Biomarker for Gastric Cancer/Xia et al Table 3. Serum MIF and CEA in Predicting Relative Risks of Various Prognostic Factors of Gastric Cancer (n¼97) Factors MIF, Cutoff 6600 pg/ml Relative Risk (95% CI) CEA, 5 ng/ml MIF1CEA* Survival <5 yvs 5 y 5.63 ( )y 3.92 ( )y 7.75 ( )z Tumor size 5 cmvs<5 cm 2.30 ( ) 2.18 ( ) 3.25 ( ) Tumor site Antrum vs other sites 0.90 ( ) 1.28 ( ) 1.70 ( ) Differentiation Poor/Moderate vs well 0.55 ( ) 0.78 ( ) 0.66 ( ) Depth of invasion T4 vs T1/T2/T ( ) 1.25 ( ) 1.81 ( ) Lymph node metastasis N1/N2 vs N ( ) 1.20 ( ) 2.08 ( ) Peritoneal/hepatic metastasis Yes vs no 2.25 ( ) 2.22 ( ) 3.36 ( ) Stage III/IV vs I/II 3.34 ( ) 2.31 ( ) 3.93 ( )y Operability Palliative/Conservative vs operable 1.78 ( ) 2.39 ( ) 4.13 (1.59 (10.73)y CI indicates confidence interval; MIF, macrophage migration-inhibitory factor; CEA, carcinoembryonic antigen. * Positive by either MIF or CEA serology. y P <.01 z P <.001 P <.05 5 years (OR ¼ 3.92; 95% CI, ; P ¼.002). In addition, serum CEA was associated with operability (Table 3). The prognostic values improved significantly when serum MIF and serum CEA were considered together; 81.2% of patients with serum MIF higher than 6600 pg/ ml and/or with serum CEA higher than 5 ng/ml and 35.7% of those with both serum MIF and CEA lower than the above levels survived less than 5 years (OR ¼ 7.75; 95% CI, ; P <.001, Table 3). Multivariate analysis showed that serum MIF, CEA, differentiation, stage, and operability were independent prognostic factors for gastric cancer (Table 4). The survival curves according to the serum MIF and CEA were shown in Figure 1. Patients with serum MIF concentrations <6600 pg/ml or CEA <5 ng/ml had significantly better survival than those who had higher serum Table 4. Multivariate Analysis by Cox Regression of Independent Prognostic Factors for Gastric Cancer (n¼97) Factors HR for Mortality (95% CI) Age 1.00 ( ).925 Sex, men vs women 0.86 ( ).559 Serum CEA, 5 vs <5 ng/ml 2.65 ( ).001 Serum MIF, 6600 vs <6600 pg/ml 2.46 ( ).001 Tumor site, antrum/incisura 1.57 ( ).127 vs. other sites Tumor size, 5 vs <5 cm 0.83 ( ).504 Differentiation, poor/moderate vs well 2.41 ( ).002 Stage, III/IV vs I/II 3.50 ( ).004 Operability, palliative/conservative vs operable 6.77 ( ) <.001 HR indicates hazard ratio; CI, confidence interval; CEA, carcinoembryonic antigen; MIF, macrophage migration-inhibitory factor. P Cancer December 1,

6 FIGURE 1. Survival curves are depicted according to serum MIF (A, cutoff value 6600 pg/ml, P ¼.0021), CEA (B, cutoff value 5 ng/ml, P ¼.0014) and a combination of serum MIF and CEA (C, P ¼.0001). levels (log-rank test, P ¼.0021 and P ¼.0014, respectively) (Fig. 1A-B). Using a combination of serum MIF higher than 6600 pg/ml and/or serum CEA higher than 5 ng/ml, the prediction for 5-year survival was even better (P ¼.0001) (Fig. 1C). DISCUSSION Previous studies have shown that MIF is overexpressed in gastric cancer cells and implicated in H. pylori-induced gastric carcinogenesis In our recent study, we observed that both serum MIF levels and gastric epithelial MIF expression were significantly increased in gastric intestinal metaplasia and gastric cancer, and the serum MIF concentration correlated well with gastric epithelial MIF expression. 22 Therefore, serum MIF level may serve as a biomarker for precancerous lesions and gastric cancer in patients with dyspepsia. In the present study, sensitivity, specificity, and accuracy of 83.5%, 92.3%, and 89.7%, respectively, were achieved for diagnosing gastric cancer with the cutoff value of 3230 pg/ml. In the meantime, we observed that serum CEA had sensitivity, specificity, and accuracy of 60.8%, 83.3%, and 76.5%, respectively, with the widely accepted cutoff value of 5 ng/ml. 23,24,33,34 This observation was consistent with previous studies that suggest that serum CEA has limited value in diagnosing gastric cancer. 33,34 Hence, serum MIF as a biomarker would increase the accuracy by about 15%, compared with serum CEA. A combination of the 2 biomarkers may improve the sensitivity and negative predictive value, but offset by decreased specificity. These findings suggest that serum MIF level may serve as a valuable biomarker for screening gastric cancer in patients presenting with dyspepsia. Dyspepsia is experienced by approximately 25%- 40% of general population. 35 Endoscopy is the most accurate method in diagnosing dyspepsia-associated conditions such as gastric cancer, peptic ulcer disease, esophagitis, and gastroduodenitis. However, endoscopy is in invasive procedure, which causes discomfort and involves cost. Therefore, identification of patients who could benefit the most from endoscopy, especially those with gastric cancer, is an important issue. It is recommended that dyspeptic patients over the age of 45, and those with certain alarm symptoms (such as anemia, hematemesis, weight loss, recurrent vomiting, dysphagia, and abdominal mass) undergo upper endoscopy, as some studies have shown that age and alarm symptoms are predictive for the major endoscopic findings including gastric cancer However, several studies failed to confirm the predictive values of age and alarm symptoms as a considerable proportion of patients with gastric cancer present themselves only 5446 Cancer December 1, 2009

7 MIF as a Biomarker for Gastric Cancer/Xia et al with uncomplicated dyspepsia Therefore, better prediction strategies are required to identify patients with gastric cancer. Serum MIF appears to be a useful biomarker to guide the physicians whether to refer the patient for prompt endoscopy, especially in areas with a moderate or high risk for gastric cancer. For example, dyspeptic patients with a high level of serum MIF should undergo upper endoscopy regardless of age and the presence of alarm symptoms, whereas young dyspeptic patients with a low level of MIF but without alarm symptoms will not need endoscopy. Serum MIF has been suggested to be of prognostic value in prostate cancer. 9 Meyer-Siegler et al reported that prostate carcinoma patients with high serum MIF levels (>6000 pg/ml) had a higher risk of recurrence compared with patients who had lower serum MIF levels. 9 In the present study, we observed that patients with serum MIF concentrations <6600 pg/ml were more likely to survive longer than those who had higher concentrations. Similar prognostic significance was also observed for serum CEA, which confirms previous studies. 23,24 Moreover, our data show that a combination of serum MIF and CEA further improves the prognostic value for survival. We have previously demonstrated that H. pyloriinduced chronic gastritis is associated with increased MIF expression by both the gastric epithelial cells and the inflammatory cells. 20,22 Moreover, it has been reported that MIF may be centrally involved in the development and progression of tumors by inhibiting apoptosis, but promoting cell proliferation, tumor angiogenesis, and metastasis through different molecular pathways. 6,14,22,41 Thus, it is conceivable that gastric cancer patients with a higher level of MIF expression would have an increased risk of progressing an advanced stage due to the proneoplastic and prometastatic effects of MIF, compared with those with a lower level. Our results, however, should be interpreted with cautions. First, like CEA or many other biomarkers for cancers, MIF is not a specific biomarker for gastric cancer because the expression of MIF was also dramatically increased in other malignancies. 6-14,21 Indeed, recent studies also demonstrated that serum MIF is also a useful biomarker for diagnosing colorectal carcinoma and malignant melanoma. 42,43 Therefore, dyspeptic patients with an extremely high serum MIF concentration but without histologically confirmed gastric cancer should also be screened for other malignancies as recent studies also indicate that serum MIF levels are an early diagnostic marker in colorectal cancer and malignant melanoma. 44,42 Likewise, our recommendation is to use MIF as a biomarker in patients presenting with dyspepsia, and not as a general screening tumor marker for asymptomatic subjects. Second, the actual incidence of gastric cancer in dyspeptic patients is much lower than the patients included in this study, and, thus, a large-scale epidemiological study may be required to draw a conclusion on the diagnostic value of serum MIF for gastric cancer. Finally, a standardized method for measuring serum MIF has not been well established, although ELISA per se is an established technique and has been used by many investigators. The reported serum MIF levels vary among different studies, with the mean serum MIF levels from 1000 to 6500 pg/ml in healthy individuals without organic diseases and from 4000 pg/ml to 35,200 pg/ml in patients with different diseases such as inflammation and tumors. 9,10,16-18,43,44 Although the difference in the mean MIF levels may reflect the finding that serum MIF concentrations are actually different among different diseases, the possibility that different ELISA methodologies applied in different studies may account for the variation in the serum MIF levels cannot be excluded. Therefore, a standard method is required for the detection of serum MIF concentrations. In conclusion, serum MIF levels may be a better biomarker than serum CEA levels in diagnosing gastric cancer in patients with dyspepsia. A combination of serum MIF and CEA predicts 5-year survival better than the individual test. Therefore, serum MIF is a potential diagnostic and prognostic biomarker for gastric cancer. Conflict of Interest Disclosures The authors made no disclosures. References 1. Murray CJL, Lopez AD. Mortality by cause for 8 regions of the world: Global burden of disease study. Lancet. 1997;349: Berrino F, Esteve J, Coleman MP. Basic issues in estimating and comparing the survival of cancer patients. IARC Scientific Publications. 1995;132:1-14. Cancer December 1,

8 3. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics: CA Cancer J Clin. 2005;55: Leung WK, Wu MS, Kakugawa Y, et al. Screening for gastric cancer in Asia: current evidence and practice. Lancet Oncol. 9: , 2008 Mar. 5. Xia HHX, Wong BCY, Lam SK. Helicobacter pylori infection and gastric cancer. Asian J Surg. 2001;24: Lue H, Kleemann R, Calandra T, Roger T, Bernhagen J. Macrophage migration inhibitory factor (MIF): mechanisms of action and role in disease. Microb Infect. 2002;4: Shimizu T, Abe R, Nakamura H, Ohkawara A, Suzuki M, Nishihira J. High expression of macrophage migration inhibitory factor in human melanoma cells and its role in tumor cell growth and angiogenesis. Biochem Biophys Res Com. 1999;264: Bin Q, Johnson BD, Schauer DW, Casper JT, Orentas RJ. Production of macrophage migration inhibitory factor by human and murine neuroblastoma. Tumour Biol. 2002;23: Meyer-Siegler KL, Bellino MA, Tannenbaum M. Macrophage migration inhibitory factor evaluation compared with prostate specific antigen as a biomarker in patients with prostate carcinoma. Cancer. 2002;94: Akbar SM, Abe M, Murakami H, et al. Macrophage migration inhibitory factor in hepatocellular carcinoma and liver cirrhosis; relevance to pathogenesis. Cancer Lett. 2001; 171: Kamimura A, Kamachi M, Nishihira J, et al. Intracellular distribution of macrophage migration inhibitory factor predicts the prognosis of patients with adenocarcinoma of the lung. Cancer. 2000;89: Bini L, Magi B, Marzocchi B, et al. Protein expression profiles in human breast ductal carcinoma and histologically normal tissue. Electrophoresis. 1997;18: Mitchell RA, Bucala R. Tumor growth-promoting properties of macrophage migration inhibitory factor (MIF). Sem Cancer Biol. 2000;10: Takahashi N, Nishihira J, Sato Y, et al. Involvement of macrophage migration inhibitory factor (MIF) in the mechanism of tumor cell growth. Mol Med. 1998;4: Maaser C, Eckmann L, Paesold G, Kim HS, Kagnoff MF. Ubiquitous production of macrophage migration inhibitory factor by human gastric and intestinal epithelium. Gastroenterology. 2002;122: Murakami H, Akbar SMF, Matsui H, Onji M. Macrophage migration inhibitory factor in the sera and at the colonic mucosa in patients with ulcerative colitis: clinical implications and pathogenic significance. Eur J Clin Invest. 2001;31: de Jong YP, Abadia-Molina AC, Satoskar AR, et al. Development of chronic colitis is dependent on the cytokine MIF. Nat Immunol. 2001;2: Ohkawara T, Nishihira J, Taeda H, et al. Amelioration of dextran sulfate-induced colitis by anti-macrophage migration inhibitory factor antibody in mice. Gastroenterology. 2002;123: Huang XR, Hui CWC, Chen YX, et al. Macrophage migration inhibitory factor is an important mediator in the pathogenesis of gastric inflammation in rats. Gastroenterology. 2001;121: Xia HHX, Lam SK, Huang XR, et al. Helicobacter pylori infection is associated with increased expression of macrophage migration inhibitory factor by epithelial cells, T-cells and macrophages in gastric mucosa. J Infect Dis. 2004;190: Xia HHX, Zhang ST, Lin MCM, Kung HF, Lam SK, Wong BCY. Expression of macrophage migration inhibitory factor in esophageal squamous cell carcinoma. Carcinogenesis. 2004;26: He XX, Yang J, Ding YW, Liu W, Shen QY, Xia HHX. Increased epithelial and serum expression of macrophage migration inhibitory factor in gastric cancer: potential role of MIF in gastric carcinogenesis. Gut. 2006; 55: Chan AOO, Lam SK, Chu KM, et al. Soluble E-cadherin is a valid prognostic marker in gastric carcinoma. Gut. 2001;48: Chan AOO, Chu KM, Lam SK, et al. Soluble E-cadherin is an independent pretherapeutic factor for long-term survival in gastric cancer. J Clin Oncol. 2003;21: Nishi M, Omori Y, Miwa K, eds. Japanese Research Society for Gastric Cancer. Japanese Classification of Gastric Carcinoma. Tokyo: Kanehara; Lauren P. The 2 histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. Acta Pathol Microbiol Scand. 1965;64: Wong BCY, Wong WM, Wang WH, et al. An evaluation of invasive and non-invasive tests for the diagnosis of Helicobacter pylori infection in Chinese. Aliment Pharmacol Ther. 2001;15: Xia HHX, Wong BCY, Wong WM, et al. Optimal serological tests for the detection of Helicobacter pylori infection in Chinese population. Aliment Pharmacol Ther. 2002; 16: Dixon MF, Genta RM, Yardley JH, Correa P. Classification and grading of gastritis. The updated Sydney system. International Workshop on the Histopathology of Gastritis, Houston Am J Surg Pathol. 1996;20: Shun CT, Lin JT, Huang SP, Lin MT, Wu MS. Expression of macrophage migration inhibitory factor is associated with enhanced angiogenesis and advanced stage in gastric carcinomas. World J Gastroenterol. 2005;11: Xia HH, Lam SK, Chan AO, et al. Macrophage migration inhibitory factor stimulated by Helicobacter pylori increases proliferation of gastric epithelial cells. World J Gastroenterol. 2005;11: Cancer December 1, 2009

9 MIF as a Biomarker for Gastric Cancer/Xia et al 32. Beswick EJ, Pinchuk IV, Suarez G, Sierra JC, Reyes VE. Helicobacter pylori CagA-dependent macrophage migration inhibitory factor produced by gastric epithelial cells binds to CD74 and stimulates procarcinogenic events. J Immunol. 2006;176: Shimizu N, Wakatsuki T, Murakami A, et al. Carcinoembryonic antigen in gastric cancer patients. Oncology. 1987;44: Wobbes T, Thomas CM, Segers MF, Nagengast FM. Evaluation of 7 tumor markers (CA50, CA19-9, CA 19-9 TruQuant, CA 72-4, CA 195, carcinoembryonic antigen, and tissue polypeptide antigen) in the pretreatment sera of patients with gastric carcinoma. Cancer. 1992;69: American Gastroenterological Association medical position statement: evaluation of dyspepsia. Gastroenterology 1998; 114: Christie J, Shepherd NA, Codling BW, Valori RM. Gastric cancer below the age of 55: implications for screening patients with uncomplicated dyspepsia. Gut. 1997;41: Canga C, Vakil N. Upper GI malignancy, uncomplicated dyspepsia, and the age threshold for early endoscopy. Am J Gastroenterol. 2002;97: Wallace MB, Durkalski VL, Vaughan J, et al. Age and alarm symptoms do not predict endoscopic findings among patients with dyspepsia: a multicentre database study. Gut. 2001;49: Maconi G, Kurihara H, Panizzo V, et al. Gastric cancer in young patients with no alarm symptoms: focus on delay in diagnosis, stage of neoplasm and survival. Scand J Gastroenterol. 2003;38: Hammer J, Eslick GD, Howell SC, Altiparmak E, Talley NJ. Diagnostic yield of alarm features in irritable bowel syndrome and functional dyspepsia. Gut. 2004;53: Bach JP, Rinn B, Meyer B, Dodel R, Bacher M. Role of MIF in inflammation and tumorigenesis. Oncology. 2008; 75: Yasasever V, Camlica H, Duranyildiz D, Oguz H, Tas F, Dalay N. Macrophage migration inhibitory factor in cancer. Cancer Invest. 2007;25: Lee H, Rhee H, Kang HJ, et al. Macrophage migration inhibitory factor may be used as an early diagnostic marker in colorectal carcinomas. Am J Clin Pathol. 2008;129: Lee H, Rhee H, Kang HJ, et al. Macrophage migration inhibitory factor my b used as an early diagnostic marker in colorectal carcinomas. Am J Clin Pathol. 2007;129: Cancer December 1,