© 2004 Foundation for Promotion of Cancer Research
Sialyl Lewisa Expression as a Predictor of the Prognosis of Colon Carcinoma Patients in a Prospective Randomized Clinical Trial
1 Department of Clinical Research, 2 Department of Gastrointestinal Surgery, Aichi Prefectural Hospital, Okazaki, Aichi, 3 Department of Surgery, Yokoyama Gastrointestinal Hospital, Nagoya, 4 Second Department of Surgery, Nagoya University, Nagoya and 5 Department of Epidemiological and Clinical Research Information Management, Kyoto University, Kyoto, Japan
For reprints and all correspondence: Takanori Matsui, Department of Clinical Research, Aichi Prefectural Hospital, 18 Kuriyado Kakemachi, Okazaki city, Aichi 444-0011, Japan. E-mail: matsui-ngy{at}umin.ac.jp
Received February 27, 2004; accepted August 16, 2004
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
Background: The metastatic potential of tumors is dependent on the cell to cell adhesion by cell surface carbohydrate antigens. Thus, expression of sialyl Lewisa, which is one of the important molecules of cell surface carbohydrates, may serve as a prognostic marker of aggressive and metastasizing tumor growth. However, the prognostic value of sialyl Lewisa expression in colon cancer is still controversial.
Methods: In this study, we investigated the expression of sialyl Lewisa antigen in 233 colon cancer specimens from patients who were registered in a prospective adjuvant immunochemotherapy clinical trial. The clinical course and the prognosis of the patients were evaluated after all the immunohistochemical analyses had been performed.
Results: Sialyl Lewisa expression levels were correlated with both overall survival (P = 0.0006) and disease-free survival (P = 0.004) in all patients with the log-rank test. This result could be assumed to have been influenced by the difference in the metastatic preponderance in a high versus low sialyl Lewisa expression in the tumor cells.
Conclusion: This prospective study in a randomized controlled trial suggests that sialyl Lewisa expression levels may serve as an indicator of the metastatic potential of colon cancer cells, which would strongly predict the prognosis.
Key Words: sialyl Lewisa • colon cancer • prognosis • prospective randomized clinical trial
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
Tumor development is often associated with changes in cell surface carbohydrates (1). The changes in cell surface carbohydrates can be classified into changes in the terminal carbohydrate structure, which include incomplete synthesis and modification of normally existing carbohydrates, and carbohydrate core structure. In some solid tumors, incomplete synthesis of the cell surface carbohydrates results in an increased expression of the precursor structures of blood-group antigens, including sialyl Lewisa. Recent studies suggest that sialyl Lewisa and other blood-group related antigens could act as cell to cell adhesion molecules (1,2) and could also be involved in tumor angiogenesis (3), which possibly influence patients' prognosis. To date, however, the definitive relation between the expressions and clinical courses has not been established in an evidence-based clinical trial.
In order to answer this question, we conducted an immunohistochemical study of sialyl Lewisa antigen expression in colon cancer patients with macroscopic lymph node metastases who were registered in a prospective randomized trial of adjuvant immunochemotherapy. In this study, the prognostic value of sialyl Lewisa expression was prospectively examined.
|
SUBJECTS AND METHODS |
|---|
|
TOPAbstractINTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
CLINICAL TRIAL
Between February 1991 and March 1993, 446 patients were enrolled in a prospective randomized trial, conducted by the ‘Study Group of Colon Cancer Immunotherapy with PSK plus 5-FU’ (CIP), to compare the efficacy of immunochemotherapy with 5-FU plus Polysaccharide K (PSK) to chemotherapy with 5-FU alone as an adjuvant therapy in patients with curatively resected colon cancer with macroscopically positive lymph node metastases (4). The therapy consisted of post-operative continuous 5-FU infusion for 4 weeks (2000 mg/m2/48 h/week) plus 18 months of oral 5-FU (200 mg/body/day), with or without altering 18 months of PSK (3 g/body/day). The 5-year disease-free survival rate was 76.8% in the immunochemotherapy group with PSK (n = 220) and 74.2% in the control chemotherapy group (n = 221). The 5-year survival rate was 83.2% in the immunochemotherapy group and 80.5% in the chemotherapy group. Detailed study design and the final 7-year results of the CIP trial are being published elsewhere.
PROSPECTIVE STUDY DESIGN FOR DETERMINING THE CORRELATION BETWEEN EXPRESSION OF SIALYL LEWISa AND OTHER BIOMARKERS AND PROGNOSIS IN PATIENTS WITH COLORECTAL CANCER
The development of monoclonal antibodies directed against sialyl Lewisa and its potential applicability in clinical trials prompted the CIP study group to design a study to examine the correlation between sialyl Lewisa expression level and the prognosis of patients registered in the trial. The CIP Genetic Marker Committee (CIP-GMC) was established from the CIP study group in April 1995. A letter was sent by the CIP-GMC to all the participating institutions requesting them to provide paraffin-embedded tissue blocks obtained from patients enrolled in the CIP trial. After obtaining clearance from the institutional review board (IRB) at each institution, 47 out of a total of 93 participating institutions agreed to provide the specimens. Thus, paraffin-embedded tissue blocks of 233 of the total 446 registered patients were obtained by the CIP-GMC. Collection of the paraffin blocks was completed in March 1996, and the blocks were cut into 4-µm thick sections at four meta-centers of the trial and sent to Aichi Prefectural Hospital, where the immunohistochemical analysis for sialyl Lewisa expression was performed. Other biomarkers were also tested by the members of the CIP-GMC. The protocol of the CIP trial was revised by the steering committee of the CIP study in July 1995. A hypothesis stating that those patients with high sialyl Lewisa expression levels and/or positive for other biomarkers had a poorer prognosis was added in a subsidiary protocol.
MONOCLONAL ANTIBODY
The monoclonal antibody used was CA19-9 (Centocor, Toray Fuji Bionics Co., Tokyo, Japan) (5), which is reactive with sialyl Lewisa. This antibody was diluted 1000 times before use.
HISTOPATHOLOGY AND IMMUNOHISTOCHEMISTRY
From August to September 1996, sections prepared from paraffin-embedded formalin-fixed tissue blocks of resected tumors were stained by immunohistochemistry. A mean of 2.5 sections per person were analyzed for sialyl Lewisa expression. The monoclonal antibody CA19-9 was used to detect sialyl Lewisa, using the standard avidin–biotin–peroxidase complex (Vectastain) technique. Briefly, tumor sections (4-µm thick) were deparaffinized in xylene and then hydrated in decreasing concentrations of ethanol. To quench the endogenous peroxidase activity, sections were incubated in a solution of 3% hydrogen peroxide for 15 min. To reduce non-specific background staining, these sections were exposed to 20% horse serum for 30 min. The sections were then incubated overnight with CA19-9 at 4°C. Thereafter, they were rinsed in PBS, incubated with biotinylated horse anti-mouse secondary antibodies, rinsed again in PBS, and then incubated with the avidin–biotin–peroxidase complex. After further rinsing, immunostaining for sialyl Lewisa was accomplished by immersing the sections in 0.05% 3,3'-diaminobenzidine tetrahydrochloride solution, followed by counterstaining with hematoxylin. The immunohistochemical staining process was completed by the end of September 1996.
SCORING OF IMMUNOHISTOCHEMICAL STAINING WITH SIALYL LEWISa
Each slide was stained and compared with positive and negative reference sections, previously classified as low or high intensity staining when using CA19-9. The intensity of the sialyl Lewisa staining of the tumor cells was arbitrarily graded from 0 to 3 (0, negative; 1, slight staining in a few cells; 2, cytoplasmic staining; 3, homogenously strong staining in the cytoplasm and in the interstitial cells). Sialyl Lewisa expression grading was performed by two independent observers, with the observer agreement over 90%. In case of disagreement, grading was determined by consensus with a third observer. Grading was completed in October 1996 and the final results of the immunohistochemical analysis were reported and deposited with the CIP-GMC in November 1996, 18 months before the start of the first analysis regarding the correlation between sialyl Lewisa expression and patient prognosis, which was started in April 1998.
STATISTICAL ANALYSIS
The Gehan Wilcoxon univariate test and multivariate analysis were used to examine the possible relationships between disease-free survival (DFS), over all survival (OS), gender, Dukes' stage, tumor site, tumor differentiation and sialyl Lewisa expression levels. When DFS was used as the endpoint, disease event referred to the recurrence of the disease, death from cancer or death from non-cancer causes. Survival curves were constructed using the Kaplan–Meier method.
|
Results |
|---|
|
TOPAbstractINTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
SIALYL LEWISa EXPRESSION
Eighty-three of the 233 patients showed high sialyl Lewisa expression levels, defined as high-intensity staining with a sialyl Lewisa score of 2 or 3. Among the patients with low-intensity sialyl Lewisa staining, 108 had a staining score of 0, and 42 had a score of 1 (Fig. 1). No significant difference was observed in sialyl Lewisa expression levels in relation to age, sex, site of primary tumor or histological type of the tumor (Table 1). However, sialyl Lewisa expression had a weak correlation with the preoperative serum CEA level and treatment groups (with or without PSK), but they were also not statistically significant (Table 1). Table 2 shows the multivariate analysis that revealed that the variables with statistical significance were Dukes' stage, pathological lymph node metastasis and sialyl Lewisa expression level.
|
|
|
OTHER BIOMARKERS OF COLON CANCER
Results of other biomarkers tested by the members of the CIP-GMC, i.e. microsatellite instability of BAT26, loss of heterozygosity at C117-503, C117-732, D17S261, immunohistochemical expression of metalloproteinase (6), CD44, laminin, pyrimidine phosphorylase, DPD and TS (7), were reported to the CIP-GMC by each investigator.
RELATIONSHIP BETWEEN SIALYL LEWISa EXPRESSION AND CLINICAL COURSE
Figures 2 and 3 show the OS and DFS of the groups classified by the immunohistochemical status. As mentioned above, we initially scored the results to four grades; however, score 0 (negative) and score 1 (slight staining in a few cells) cases showed almost the same prognosis, and other clinical factors also appeared to be similar. Therefore, further analyses were carried out by comparing two groups (score 0 and 1 as low sialyl Lewisa group and score 2 and 3 as high sialyl Lewisa group).
|
|
Sialyl Lewisa expression levels were inversely correlated with both OS (P = 0.0006) and DFS (P = 0.004) in all the patients with the log-rank test. OS and DFS at 5 years were 71.1% and 70.8%, respectively, in sialyl Lewisa high group and 88.0% and 86.5%, respectively, in sialyl Lewisa low group (Figs 2 and 3). Thus, among patients with low sialyl Lewisa expression levels, an additional 15.7% were disease-free and an additional 16.9% were alive after 5 years compared with patients with high sialyl Lewisa expression levels. Table 3 shows the relation between recurrences and sialyl Lewisa expression levels. Liver metastasis, which is mostly observed as a recurrence after curative surgery of colon cancer, occurred 1.6 times more frequently in the sialyl Lewisa high group compared with the sialyl Lewisa low group, although it was not statistically significant. Metastases to the lung, lymph nodes and the ovary were observed significantly more frequently in the sialyl Lewisa high group compared with the sialyl Lewisa low group (Table 3). This higher incidence of metastases in the sialyl Lewisa high group could be an important factor that influences prognosis.
|
|
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
Sialyl Lewisa, also known as CA19-9, is one of the blood group-related antigens that is expressed in malignant tumor cells including colon cancer. Recently, the function of this molecule has been getting clarified by various in vitro studies. Takada et al. reported that sialyl Lewisa also serves as a ligand for endothelial leukocyte adhesion molecule-1 (ELAM-1) (2). This may help tumor cells to adhere to the vessels, which is thought to be the initial step in metastasis. Sialyl Lewisa may also act as a tumor angiogenesis factor (3). This may assist rapid growth of both primary and metastatic tumor, resulting in poor prognosis.
Considering these in vitro results, the expression of sialyl Lewisa might lead to poor prognosis for the patients, in terms of both, increasing the incidence of metastasis and faster progression of the tumor. Several investigators have tried to prove this hypothesis in a variety of clinical settings, but until now, no conclusive evidence has emerged (8–15). Shimono et al. (n = 149) (8), Nakayama et al. (n = 309) (9) and Asai et al. (n = 308) (12) reported that sialyl Lewisa antigen expression in colorectal tumors is a useful marker for evaluating tumor aggressiveness and the prognosis of the patients. On the other hand, Nakagoe et al. (n = 101) (11) and Arends et al. (n = 311) (14) reported that the presence of sialyl Lewisa in gastrointestinal cancers is not correlated with the parameters known to be of prognostic significance. Walker and Day [n = 37 (only those cases in which pathological findings were reported)] (15) found no correlation between the presence of sialyl Lewisa in breast cancer and lymph node status. Thus, the relation between sialyl Lewisa expression and metastatic behavior remains unclear.
Shimono et al. (8), on reviewing their references, mentioned that this discrepancy might be the result of the different classifications of the staining pattern and the smaller number of examined cases. The positive ratio of Shimono's study was 58%, while those of other studies were 54–90%, most of which were higher than in Shimono's study. In our study, the ratio of positively stained cases was 35.6% (83/233); thus, we agree with Shimono's opinion that the prognosis of the faintly stained cases was similar to that of negative cases.
In addition to the above, we think that the difference in the results may, at least partly, be due to poorly designed retrospective protocols. Recently, a meta-analysis showed the efficacy of adjuvant oral fluoropyrimidines (16). However, before accepting this as evidence, it should be determined whether the patient received post-operative adjuvant chemotherapy since the patient's attitude or other factors could be different in each case. Therefore, the conditions of adjuvant chemotherapies have varied in retrospective studies, which make it difficult to analyze the data properly. Perhaps, because of this, none of the referred studies mentioned the post-operative adjuvant chemotherapy, which could play an important role in the prognosis. In our study, the adjuvant chemotherapy received by the patients was invariant in a prospective setting, which could give us more important information on the sialyl Lewisa regardless of the conditions of the adjuvant chemotherapy.
Our results also show the higher incidence of the metastases and, consequently, poor prognosis, in sialyl Lewisa positive cases, which is consistent with the in vitro studies. In this study, we revealed a significant correlation between high sialyl Lewisa expression group and high incidence of metastases, which consequently decreases the overall survival and disease-free survival rates. However, with regard to the liver, which is one of the most important sites to evaluate the recurrence after resection of colon cancer, statistically significant results were not obtained. This suggests that additional factors play other unknown important roles in liver metastasis. Although it was not statistically significant, probably because of the lack of enough events, the rate of liver metastasis was higher in the group with high sialyl Lewisa than in those with low sialyl Lewisa. Thus, the possible importance of sialyl Lewisa expression related to liver recurrence can still not be excluded.
All cases analyzed in this study were treated by 5-FU-based adjuvant therapies, and the results may be relevant only in this special clinical setting. From our results, the effects of adjuvant therapy on both high and low sialyl Lewisa groups are still unknown. To answer this, further investigation that includes ‘no chemotherapy’ arm would be required. However, now that evidence that survival rate is better in adjuvant chemotherapy groups than in no chemotherapy groups has been established (16), it has become very difficult to plan prospective investigations with such an arm.
In conclusion, sialyl Lewisa could be one of the independent useful prognostic markers for colon cancer patients after curative resection. The results suggest that this is due to the difference in the metastatic potential of the cancer cells, which has been shown by in vitro studies.
|
References |
|---|
|
TOPAbstractINTRODUCTIONSUBJECTS AND METHODSResultsDISCUSSIONReferences |
|---|
1 Dabelsteen E. Cell surface carbohydrates as prognostic markers in human carcinomas. J Pathol 1996;179:358–69.[CrossRef][Web of Science][Medline]
2 Takada A, Ohmori K, Yoneda T, Tsuyuoka K, Hasegawa A, Kiso M, et al. Contribution of carbohydrate antigens sialyl Lewis A and sialyl LewisX to adhesion of human cancer cells to vascular endothelium. Cancer Res 1993;53:354–61.
3 Tei K, Kawakami-Kimura N, Taguchi O, Kumamoto K, Higashiyama S, Taniguchi N, et al. Roles of cell adhesion molecules in tumor angiogenesis induced by cotransplantation of cancer and endothelial cells to nude rats. Cancer Res 2002;62:6289–96.
4 Nakazato H et al., Proc ASCO 18: abs. 245, 1999.
5 Koprowski H, Steplewski Z, Mitchell K, Herlyn M, Herlyn D, Fuhrer P. Colorectal carcinoma antigens detected by hybridoma antibodies. Somatic Cell Genet 1979;5:957–71.[CrossRef][Web of Science][Medline]
6 Ougolkov AV, Yamashita K, Mai M, Minamoto T. Oncogenic beta-catenin and MMP-7 (matrilysin) cosegregate in late-stage clinical colon cancer. Gastroenterology 2002;122:60–71.[CrossRef][Web of Science][Medline]
7 Sakamoto J, Hamashima H, Suzuki H, Ito K, Mai M, Saji S, et al. Thymidylate synthase expression as a predictor of the prognosis of curatively resected colon carcinoma in patients registered in an adjuvant immunochemotherapy clinical trial. Oncol Rep 2003;10:1081–90.[Medline]
8 Shimono R, Mori M, Akazawa K, Adachi Y, Sugimachi K. Immunohistochemical expression of carbohydrate antigen 19-9 in colorectal carcinoma. Am J Gastroenterol 1994;89:101–5.[Medline]
9 Nakayama T, Watanabe M, Katsumata T, Teramoto T, Kitajima M. Expression of sialyl Lewis A as a new prognostic factor for patients with advanced colorectal carcinoma. Cancer 1995;75:2051–6.[CrossRef][Medline]
10 Ikeda Y, Mori M, Kajiyama K, Haraguchi Y, Sasaki O, Sugimachi K. Immunohistochemical expression of sialyl Tn, sialyl Lewis a, sialyl Lewis a-b-, and sialyl LewisX in primary tumor and metastatic lymph nodes in human gastric cancer. J Surg Oncol 1996;62:171–6.[CrossRef][Medline]
11 Nakagoe T, Fukushima K, Nanashima A, Sawai T, Tsuji T, Jibiki M, et al. Expression of Lewis a, sialyl Lewis a, LewisX and sialyl LewisX antigens as prognostic factors in patients with colorectal cancer. Can J Gastroenterol 2000;14:753–60.[Medline]
12 Asai S, Watanabe T, Sakamoto J, Kiriyama K, Ito K, Akiyama S, et al. Expression and prognostic indicators of type 1 and type 2 Lewis blood group antigens in colorectal cancers. Nippon Geka Gakkai Zasshi 1994;95:753–62.[Medline]
13 Tsumatori G, Ozeki Y, Takagi K, Ogata T, Tanaka S. Relation between the serum E-selectin level and the survival rate of patients with resected non-small cell lung cancers. Jpn J Cancer Res 1999;90:301–7.[CrossRef][Web of Science]
14 Arends JW, Wiggers T, Verstijnen C, Hilgers J, Bosman FT. Gastrointestinal cancer-associated antigen (GICA) immunoreactivity in colorectal carcinoma in relation to patient survival. Int J Cancer 1984;34:193–6.[Medline]
15 Walker RA, Day SJ. Expression of the antigen detected by the monoclonal antibody CA19-9 in human breast tissues. Virchows Arch A Pathol Anat Histopathol 1986;409:375–83.[Medline]
16 Meta-analysis group of the Japanese Society for Cancer of the Colon and Rectum and the Meta-analysis Group in Cancer. Efficacy of oral adjuvant therapy after resection of colorectal cancer: 5-year results from three randomized trials. J Clin Oncol 2004;22:484–92.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Malagolini, D. Santini, M. Chiricolo, and F. Dall'Olio Biosynthesis and expression of the Sda and sialyl Lewis x antigens in normal and cancer colon Glycobiology, July 1, 2007; 17(7): 688 - 697. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||