ANTICANCER RESEARCH 26: (2006)

Transcription

1 Expression of Tn and Sialyl-Tn Antigens in Endometrial Cancer: Its Relationship with Tumor-produced Cyclooxygenase-2, Tumor-infiltrated Lymphocytes and Patient Prognosis SATOSHI OHNO 1,2, YUMIKO OHNO 2, HIROSHI NAKADA 3, NOBUTAKA SUZUKI 1,2, GEN-ICHIRO SOMA 4,5 and MASAKI INOUE 2 1 Department of Complementary and Alternative Medicine and 2 Department of Obstetrics and Gynecology, Kanazawa University, Graduate School of Medical Science, Ishikawa; 3 Department of Biotechnology, Faculty of Engineering, Kyoto Sangyo University, Kyoto; 4 Institute for Health Sciences, Tokushima Bunri University, Tokushima; 5 Institute for Drug Delivery System, Tokyo University of Science, Chiba, Japan Abstract. Background: Although many tumors arising from epithelial tissues produce mucins, little is known regarding the biological significance of mucins in cancer. Recent researches have revealed that mucins produced by cancer cells play a critical role in the initial induction of cyclooxygenase (COX)-2 in the tumor microenvironment. Moreover, the consequence of abnormal glycosylation and overexpression of mucins includes suppression of the host-immune function. Materials and Methods: The expression of Tn/sialyl-Tn (s-tn) antigen and its impact on COX-2 expression by tumor cells and infiltration of CD8+ T-cell into the cancer cell nest (nest CD8) in endometrial cancer was evaluated. Tissue specimens from 70 endometrial cancer patients who had undergone a curative resection were evaluated for Tn/s-Tn antigen, COX-2 and CD8 by immunohistochemistry. Results: The overexpression of Tn and s-tn antigen was significantly correlated with COX-2 overexpression (p<0.005 and p<0.001, respectively). There was a significant association between s-tn overexpression and low infiltration of nest CD8 (p<0.05). Strong expression of s-tn antigen was significantly associated with poor prognosis (p<0.005). Among several prognostic factors, s-tn expression remained the strong independent predictor of survival by multivariate analysis (p<0.05). Conclusion: These results suggested that tumor-produced mucins, which were associated Correspondence to: Satoshi Ohno, Department of Complementary and Alternative Medicine, Kanazawa University, Graduate School of Medical Science, 13-1 Takaramachi, Kanazawa, Ishikawa, , Japan. Tel: , Fax: , satoshio@med.kanazawa-u.ac.jp Key Words: Mucin, Tn antigen, sialyl-tn antigen, cyclooxygenase-2, CD8+ T-cells, endometrial cancer, prognosis. with COX-2 induction and immunosuppression, provided additional prognostic information in endometrial cancer patients. Many tumors arising from epithelial tissues produce mucins. They are defined by their characteristic O-glycosylated domains, which contain a repetitive protein backbone with an especially high content of Ser and Thr residues. The first step in O-linked oligosaccharide synthesis is the addition of N-acetylgalactosamine to Ser or Thr residues resulting in the formation of the Tn antigen. The addition of sialic acid by an 2,6-sialosyltransferase forms the sialyl-tn (s-tn) antigen. Upon malignant transformation, many epithelial cells produce mucins in abnormal amounts and/or with an abnormal glycosylation pattern (1). Moreover, mucins produced by cancer cells are found in the sera of cancer patients and used as disease markers. S-Tn antigen is accepted as a clinically significant tumor marker for ovarian, cervical, gastric and pancreatic cancer. Moreover, Tn and s- Tn antigens have infrequently been demonstrated in normal nonsecretory endometria, and have shown an increased expression in adenomatous hyperplasias and endometrial carcinomas (2, 3). However, little is known regarding the biological significance of mucins in cancer. Recently, Inaba et al. have revealed that mucins produced by colon cancer cells play a critical role in the initial induction of cyclooxygenase (COX)-2 in the tumor microenvironment (4). COX-2, the inducible enzyme expressed at sites of inflammation, has recently come under the spotlight for its pathophysiological significance in various types of cancer. The overexpression of COX-2 is linked to angiogenesis, tumor-cell apoptosis, tumor-cell growth and metastasis in cancer progression (5). Moreover, it has been reported that tumor-produced COX-2, which reduces the infiltration of /2006 $

2 CD8+ T-cells into cancer cell nests (nest CD8), may allow tumors to avoid immune surveillance (6). In uterine endometria, COX-2 plays a critical role not only in the maintenance of the endometrium during the menstrual cycle, but also in the progression of endometrial cancer. Regarding the impact of mucin itself on host-immune status, the consequence of abnormal glycosylation and overexpression of mucins includes alteration of the adhesive capabilities of tumor cells and suppression of the function of T-cells and antigen presenting cells (7-10). Although the potential for immunosuppression associated with tumorproduced mucin has been suggested, few reports are available on the effect of mucin as an immunomodulator in the intratumoral microenvironment of human solid tumors. To date, no direct comparison has been made of the relationship between mucins produced by cancer cells and COX-2 expression and tumor-infiltrated lymphocytes (TILs) in the tumor microenvironment of endometrial cancer. To address these questions, the aim of the present work was to investigate the role of mucins in expression of COX-2 by tumor cells and immunomodulation by assessment of nest CD8 counts in surgically-removed endometrial carcinomas. This paper also discusses the prognostic significance of Tn and s-tn antigen in endometrial carcinoma patients. Materials and Methods Patients. This study included 70 patients with primary endometrial carcinoma who were consecutively admitted, treated and followedup at the Department of Obstetrics and Gynecology, Kanazawa University Hospital, Japan, from January 1995 to December All patients underwent a total abdominal or radical hysterectomy plus bilateral salphingo-oophorectomy. At the time of celiotomy, peritoneal fluid samples were obtained for cytological testing. Systemic pelvic lymphadenectomy was performed in 51 (72.9%) patients. Paraaortic lymph node sampling was performed in two patients because of visible or palpable enlarged lymph nodes. All patients were classified by the International Federation of Gynecology and Obstetrics (FIGO) surgical staging system (1988). No patients had remaining macroscopic tumors or known distant metastasis immediately after surgery. High-risk patients (e.g., deep myometrial invasion, cervical involvement, special histology, peritoneal cytology) underwent external radiotherapy and/or six cycles of chemotherapy [Paclitaxel: 180 mg/m 2, Carboplatin: according to Chatelut s formula (AUC=5 mgñmin/ml)] as postoperative adjuvant therapy. Patient treatment was followed with a gynecologic examination, recording of laboratory data, transvaginal/abdominopelvic ultrasonography, and a radiologic investigation. Data from regular follow-up visits to the outpatient department were stored in a database specifically designed for endometrial carcinoma patients. A telephone inquiry to update the present status of all surviving patients was made in July The exact date of disease recurrence was obtained from the referring physicians or from the physicians who attended the patient for the initial diagnosis of the recurrence. All treatments and clinical research were conducted with written informed consent. Immunohistochemistry. Representative paraffin sections containing both the normal endometrium and the invasive front of the tumor tissue were selected for immunohistochemical staining. Slides were deparaffinized and rehydrated in graded alcohols. Endogenous peroxidase activity was quenched by dipping in 3% hydrogen peroxide for 30 min. Nonspecific staining was blocked by treating the slides with normal serum for 30 min. The slides were incubated with mice monoclonal antibodies (MLS128 against Tn and MLS132 against s-tn, respectively) at a concentration of 1 Ìg/ml for 16 h at 4ÆC (11, 12). The subsequent steps were carried out according to the manufacturer s instructions by the HISTOFINE SAB-PO (M) kit (Nichirei Co., Tokyo, Japan). Color development was carried out with peroxidase substrate 3-amino-9-ethylcarbazole (AEC). All slides were counterstained with Mayer s hematoxylin. Sections without primary antibodies, as well as those with non-immunized mice serum, served as negative controls. COX-2 expression (tumor cells) and CD8 expression (cytotoxic T-lymphocytes) were evaluated in serial sections. The slides were incubated with mouse monoclonal antibodies against COX-2 (TaKaRa BIO Inc., Shiga, Japan; clone, CX229; dilution 1:250; overnight at 4ÆC) and CD8 (DAKO Co., Glostrup, Denmark; clone, C8/144B; dilution 1:1; for 1 hr at room temperature) as described previously (6). Evaluation of staining. In assessment of Tn and s-tn expression, the extent of immunohistochemical staining within the carcinomas was categorized as follows: negative, focally positive (less than 20% of carcinoma cells stained) or diffusely positive (more than 20% of carcinoma cells stained) (11, 12). In this study, tumors with diffusely stained cancer cells were considered to exhibit strong expression. For evaluation of COX-2, staining intensity was scored as 0 (negative), 1 (weak), 2 (medium) and 3 (strong). Extent of staining was scored as 0 (0%), 1 (1-25%), 2 (26-50%), 3 (51-75%) and 4 (76-100%) according to the percentage of positive staining area in relation to the whole carcinoma area. The sum of the intensity and extent score was used as the final staining score (0-7) for COX-2. Tumors having a final staining score of 4 were considered to exhibit strong expression. Lymphocytes stained by anti CD8 antibody within the cancer cell nests were recorded as nest CD8. For counting nest CD8, the sections were first evaluated at x100 magnification under a light microscope. Then, five representative areas within cancer cell nests where CD8+T-cells accumulated at great density were identified. CD8+T-cells were next counted at x400 magnification and the average of the five values was used as the nest CD8. Evaluation and counting were conducted by two observers (S. O. and Y. O.) who did not know any details regarding the patients' background. Statistical analysis. The chi-square test for 2x2 tables was used to compare the categorical data. In the analysis of relapse-free survival rates, those who died of causes unrelated to endometrial cancer and those who had no detected evidence of disease recurrence were considered to be relapse-free. Life tables were computed using the Kaplan-Meier method, while the log-rank test was used to assess statistical significance. Cox proportional hazards analysis was used to determine the relative contribution of various factors to the risk of recurrence. A p-value of <0.05 was considered to indicate statistical significance. All statistical analyses were performed using the statistical package StatView version 5.0 for Macintosh (Abacus Concepts, Berkeley, CA, USA). 4048

3 Ohno et al: Role of Mucins in Endometrial Cancer Results Patients characteristics. The patients' average age at the time of surgery was 57.3 years of age (range, 26-78). Patients with endometrial cancer included: 22 with pre-menopausal status, 4 with peri-menopausal status and 44 with post-menopausal status. The patients' mean pre-operative body mass index (BMI) was 24.0 (range, ). Among the 70 patients, 12 patients (17.1%) had relapses of endometrial cancer at the time of the last follow-up. The median follow-up time for all patients was 3.28 years (range, years). Tn and s-tn expression. The monoclonal antibodies against Tn and s-tn antigen reacted with the apical surface, the whole plasma membrane, the supranuclear areas (Golgi area) and the whole cytoplasm (Figure 1A-B) of endometrial cancer glands. Tn and s-tn antigens were rarely detected in the normal nonsecretory endometria. The expression of Tn antigen within carcinomas was strong (diffusely positive) in 36 patients (51%) (Figure 1A) and weak (focally positive: 10 cases; negative: 24 cases) in 34 patients (49%). The association between Tn expression and clinicopathological variables is shown in Table I. In evaluation of s-tn antigen expression, 31 patients (44%) showed strong expression (diffusely positive) (Figure 1B), whereas 39 (56%) patients demonstrated weak expression (focally positive: 12 cases; negative: 27 cases). The association between s-tn expression and clinicopathological variables is shown in Table II. COX-2 expression. The expression of COX-2 was strong (final staining score of 4-7) in 37 patients (53%) and weak (final staining score of 0-3) in 33 patients (47%). The typical COX- 2 expression in endometrial cancer cells is shown in Figure 1C. Although COX-2 overexpression was associated with advanced FIGO stage and more than half myometrial invasion, no statistically significant impact was shown (p=0.06 and p=0.09, respectively). Patient age, lymph node metastasis, grade of differentiation, menopause and BMI were not associated with COX-2 expression (data not shown). CD8 expression. CD8+ T-cells were detected within the cancer cell nest in 62 out of 70 patients (Figure 1D). The nest CD8 numbers varied from 0.0 to 59.8 with a median value of 4.1 (average±standard deviation, 10.3±14.5). All clinicopathological characteristics (patient age, FIGO stage, lymph node metastasis, myometrial invasion, grade of differentiation, menopause and BMI) were not associated with the nest CD8 number (data not shown). Associations between Tn / s-tn expression and COX-2 expression and nest CD8. The median values of nest CD8 were used as cut-off points to separate the patients into two groups (high nest CD8; 35 cases, low nest CD8; 35 cases). Strong expression of COX-2 in tumor cells was associated with strong expression of Tn and s-tn antigen (Table IIIA-B). Although an inverse association was found between expression of s-tn antigen and nest CD8, there was no correlation between expression of Tn antigen and nest CD8 (Table IIIA-B) Survival analysis. Strong expression of s-tn antigen was the factor negatively influencing the relapse-free survival rate by univariate analysis (Figure 2B). The expression of Tn antigen had no statistically significant impact on relapse-free survival (Figure 2A). After adjusting for several prognostic factors (FIGO stage, myometrial invasion, histology and s-tn antigen), strong expression of s-tn antigen and advanced FIGO stage were identified as independent predictive factors of patient survival in a Cox proportional-hazard model (Table IV). Discussion Three main points emerged from this study: 1) the overexpression of Tn and s-tn antigen was significantly correlated with COX-2 overexpression in endometrial cancer; 2) there was a significant association between s-tn overexpression and low infiltration of nest CD8; 3) strong expression of s-tn antigen was significantly associated with poor prognosis of the endometrial cancer patients. Among several prognostic factors, s-tn expression was an independent predictor of survival by multivariate analysis. Several possible explanations could account for the correlation between COX-2 expression and mucin secretion. Jang et al. reported the increased expression of COX-2 in MUC2-positive area of columnar-lined esophagus by immunohistochemistry (13). Since COX-2 is a prostaglandin (PG) synthetase, the most obvious consequence of COX-2 overexpression is increased PG production. Kim et al. pointed out that PGE 2 induced by COX-2 might play a role in mucin secretion from the gallbladder epithelium mediated by camp (14). Belley et al. discovered that PGE 2 coupling to the EP4 receptor stimulated camp-dependent mucin exocytosis in colonic epithelial cell line (15). Moreover, Kim et al. demonstrated that in human pulmonary epithelial cells, interleukin-1 activated extracellular signal-regulated kinase or p38 to induce COX-2 production, which in turn induced mucin genes (MUC2 and MUC5AC) production (16). In the tumor microenvironment, Inaba et al. revealed that mucins secreted by colon cancer cells induce production of COX-2 in tumor-infiltrating monocytes/macrophages, resulting in PGE 2 production (4). Moreover, they have proposed that the PGE 2 secreted from the macrophages binds to EP2 receptor present on cancer cells, and can work as a mediator in a paracrine fashion, leading to COX-2 production by the cancer cells. 4049

4 Figure 1. A-D: Representative sections of endometrial cancer with immunohistochemical staining of Tn antigen (A, x100), sialyl-tn antigen (B, x100), COX-2 (C, x100) and CD8 (D, x200) are shown. Accordingly, these results reported by Inaba et al. suggests that there is a strong positively correlation between mucin secretion and COX-2 expression even in the tumor microenvironment. The direct association between Tn and s-tn antigen and COX-2 expression in endometrial cancer has also been demonstrated. Further studies are needed to clarify causal relationship between mucin secretion and COX-2 production in cancer cells. In the tumor microenvironment, the consequence of abnormal glycosylation and overexpression of mucins includes alteration of the adhesive capabilities of tumor cells and suppression of the function of T-cells and in vitro generation of dendritic cells (DCs). Van de Wiel-van Kemenade et al. demonstrated that melanoma cells transfected with the MUC1 cdna-encoding episialin were significantly less susceptible to lysis than episialin-negative (non-transfected) melanoma cells (A375) by activated cytotoxic effector cells (7). Agrawal et al. revealed that cancer-associated MUC1 mucin, affinity-purified from ascite fluid of cancer patients, and synthetic tandem repeats of MUC1 mucin core peptide suppressed human T-cell proliferative responses (8). They also pointed out that this MUC1 mucin-induced suppression of T-cell responses can be reversed by the addition of exogenous IL-2 or anti-cd28 monoclonal antibody. Monti et al. investigated whether tumor-derived mucin affected the cytokine repertoire of monocyte-derived DCs (9), showing that tumor-derived mucin induced IL-10 (high) and IL-12 (low) in regulatory DC with a limited capacity to trigger protective Th1 responses. Carlos et al. also reported that human tumor antigen MUC1 was chemotactic for immature DCs, but did not promote Th1 type immunity (10). Monti et al. and Carlos et al. have provided evidence that tumor-derived mucins are responsible for impaired DC maturation and function. These reports identify mucin secretion as a novel mechanism of tumor escape from immune surveillance. In our study, an inverse correlation between s-tn expression and the infiltration of CD8+ T-cells into cancer cell nests 4050

5 Ohno et al: Role of Mucins in Endometrial Cancer Table I. Tn expression and clinicopathological characteristics. Variables Tn expression P-value (n=36) (n=34) Age (yr) <65 (n=43) (n=27) FIGO stage I (n=52) II, III, IV (n=18) 11 7 Lymph node metastasis negative (n=65) positive (n=5) 4 1 Depth (myometrial invasion) a, b (n=53) c (n=17) 9 8 Histopathology-degree of differentiation Grade 1 (n=38) Grade 2, 3, others (n=32) Menopause peri, pre (n=26) post (n=44) Body mass index <25 (n=45) (n=25) Table II. Sialyl-Tn expression and clinicopathological characteristics. Variables Sialyl-Tn expression P-value (n=31) (n=39) Age (yr) <65 (n=43) (n=27) 18 9 FIGO stage I (n=52) II, III, IV (n=18) 11 7 Lymph node metastasis negative (n=65) positive (n=5) 2 3 Depth (myometrial invasion) a, b (n=53) c (n=17) 10 7 Histopathology-degree of differentiation Grade 1 (n=38) Grade 2, 3, others (n=32) Menopause peri, pre (n=26) post (n=44) Body mass index <25 (n=45) (n=25) was discovered. Moreover, it was previously reported that COX-2 produced by tumor cells might be one of the several limiting factors controlling the degree of CD8+ T-cell infiltration into the cancer cell nest (6). Our results, which showed direct association between Tn/s-Tn antigen and COX-2 expression, may indicate that tumor-derived mucins and COX-2 work in conjunction to suppress the host immune surveillance. Although mucins produced by the secretory epithelial cells essentially work as lubrication and protection for ducts and lumina of the human body, excessive mucin secretion is a hallmark of pathogenesis in a variety of disease-related situations. To date, several authors have confirmed that the overexpression of s-tn antigen is a prognostic factor in patients with colorectal (17), gastric (18), ovarian (19), esophageal (20) and breast cancer (21). In our study, strong expression of s-tn antigen has emerged as an independent prognostic factor in endometrial cancer (Hazard ratio: 5.864, p-value: ). As established by a seminal study by Bresalier et al. (22), increased sialylation of mucinassociated carbohydrates is characteristic of colon cancer cells that are most likely to metastasize. Moreover, they have revealed that sialylated carbohydrate structures in mucin play a role in adhesive interactions involving both basement membranes and endothelial-associated ligands. In conclusion, our study has shown that tumor-produced mucins, which were associated with COX-2 induction and immunosuppression, provided additional prognostic information in endometrial cancer patients. Acknowledgements This work was supported by a Grant-in-Aid for Young Scientists (B) (No and No ) from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and by the Megumi Medical Foundation, Kanazawa, Japan. We are grateful to the staff at the Pathology Section, Kanazawa University Hospital, for collecting the samples and providing paraffinembedded tissues. References 1 Kim YS, Gum J Jr and Brockhausen I: Mucin glycoproteins in neoplasia. Glycoconj J 13: , Ravn V, Mandel U, Svenstrup B and Dabelsteen E: Simple mucin-type carbohydrates in normal and malignant human endometrium. Int J Gynecol Pathol 14: , Semczuk A, Paszkowska A, Miturski R, Skomra D, Cybulski M, Jakowicki JA and Berbec H: Sialosyl-Tn expression in normal and pathological conditions of human endometrium. An immunohistochemical study. Pathol Res Pract 198: ,

6 Table IIIA. Association between Tn expression and COX-2 expression and CD8+ T-cells infiltration. Variables Tn expression P-value (n=36) (n=34) COX-2 expression strong (n=37) week (n=33) Nest CD8* high (n=35) low (n=35) *cut-off point (nest CD8) median value (4.1). Table IIIB. Association between sialyl-tn expression and COX-2 expression and CD8+ T-cells infiltration. Variables Sialyl-Tn expression P-value (n=31) (n=39) COX-2 expression strong (n=37) week (n=33) 7 26 Nest CD8* high (n=35) low (n=35) *cut-off point (nest CD8) median value (4.1). Table IV. Multivariate analysis of prognostic factors. Cox proportional hazards model Variables Risk Hazard 95% confidence P-value factor ratio interval Figure 2. A-B: The Kaplan-Meier survival curves of 70 patients with endometrial carcinoma in relation to Tn (A) and sialyl-tn (B) antigen expression are shown. FIGO stage II Myometrial invasion >1/ Grade (Histology) Sialyl-Tn expression strong Inaba T, Sano H, Kawahito Y, Hla T, Akita K, Toda M, Yamashina I, Inoue M and Nakada H: Induction of cyclooxygenase-2 in monocyte/macrophage by mucins secreted from colon cancer cells. Proc Natl Acad Sci USA 100: , Ohno S, Ohno Y, Suzuki N, Inagawa H, Kohchi C, Soma G and Inoue M: Multiple roles of cyclooxygenase-2 in endometrial cancer. Anticancer Res 25: , Ohno Y, Ohno S, Suzuki N, Kamei T, Inagawa H, Soma G and Inoue M: Role of cyclooxygenase-2 in immunomodulation and prognosis of endometrial carcinoma. Int J Cancer 114: , van de Wiel-van Kemenade E, Ligtenberg MJL, de Boer AJ, Buijs F, Vos HL, Melief CJM, Hilkens J and Figdor CG: Episialin (MUC1) inhibits cytotoxic lymphocyte-target cell interaction. J Immunol 151: , Agrawal B, Krantz MJ, Reddish MA and Longenecker BM: Cancer-associated MUC1 mucin inhibits human T-cell proliferation, which is reversible by IL-2. Nat Med 4: 43-49, Monti P, Leone BE, Zerbi A, Balzano G, Cainarca S, Sordi V, Pontillo M, Mercalli A, Di Carlo V, Allavena P and Piemonti L: Tumor-derived MUC1 mucins interact with differentiating monocytes and induce IL-10highIL-12low regulatory dendritic cell. J Immunol 172: ,

7 Ohno et al: Role of Mucins in Endometrial Cancer 10 Carlos CA, Dong HF, Howard OM, Oppenheim JJ, Hanisch FG and Finn OJ: Human tumor antigen MUC1 is chemotactic for immature dendritic cells and elicits maturation but does not promote Th1 type immunity. J Immunol 175: , Ohshio G, Imamura T, Imamura M, Yamabe H, Sakahara H, Nakada H and Yamashina I: Distribution of Tn antigen recognized by an anti-tn monoclonal antibody (MLS128) in normal and malignant tissues of the digestive tract. J Cancer Res Clin Oncol 121: , Ohshio G, Yoshioka H, Manabe T, Sakahara H, Yamabe H, Imamura M, Inoue M, Tanaka N, Nakada H and Yamashina I: Expression of sialosyl-tn antigen (monoclonal antibody MLS102 reactive) in normal tissues and malignant tumors of the digestive tract. J Cancer Res Clin Oncol 120: , Jang TJ and Cho MY: Cyclooxygenase-2 expression and cell proliferation are increased in MUC2-positive area of columnarlined esophagus. Pathol Int 55: , Kim HJ, Lee SK, Kim MH, Seo DW and Min YI: Cyclooxygenase-2 mediates mucin secretion from epithelial cells of lipopolysaccharide-treated canine gallbladder. Dig Dis Sci 48: , Belley A and Chadee K: Prostaglandin E(2) stimulates rat and human colonic mucin exocytosis via the EP(4) receptor. Gastroenterology 117: , Kim YD, Kwon EJ, Park DW, Song SY, Yoon SK and Baek SH: Interleukin-1beta induces MUC2 and MUC5AC synthesis through cyclooxygenase-2 in NCI-H292 cells. Mol Pharmacol 62: , Itzkowitz SH, Bloom EJ, Kokal WA, Modin G, Hakomori S and Kim YS: Sialosyl-Tn. A novel mucin antigen associated with prognosis in colorectal cancer patients. Cancer 66: , Victorzon M, Nordling S, Nilsson O, Roberts PJ and Haglund C: Sialyl Tn antigen is an independent predictor of outcome in patients with gastric cancer. Int J Cancer 65: , Ghazizadeh M, Ogawa H, Sasaki Y, Araki T and Aihara K: Mucin carbohydrate antigens (T, Tn, and sialyl-tn) in human ovarian carcinomas: relationship with histopathology and prognosis. Hum Pathol 28: , Flucke U, Zirbes TK, Schroder W, Monig SP, Koch V, Schmitz K, Thiele J, Dienes HP, Holscher AH and Baldus SE: Expression of mucin-associated carbohydrate core antigens in esophageal squamous cell carcinomas. Anticancer Res 21: , Leivonen M, Nordling S, Lundin J, von Boguslawski K and Haglund C: STn and prognosis in breast cancer. Oncology 61: , Bresalier RS, Ho SB, Schoeppner HL, Kim YS, Sleisenger MH, Brodt P and Byrd JC: Enhanced sialylation of mucin-associated carbohydrate structures in human colon cancer metastasis. Gastroenterology 110: , Received May 23, 2006 Revised July 25, 2006 Accepted August 21,