Pancreatic cancer has a very poor prognosis and is one of the most common malignancies.1 Tumor resection is performed in 9–36% of patients and the 5-year survival rate of patients who have undergone resection is 19–24%.2, 3 Many attempts have been made to improve the outcome of patients' survival. Recently, significant improvements in survival have been reported compared to the survival expected from series in the 1980s.4, 5, 6, 7, 8, 9 The mechanisms responsible for the aggressiveness of this disease are not completely understood, therefore, it is very important to find novel molecular markers that are expressed frequently and predominantly in pancreatic cancer. Furthermore, the identification of factors that significantly correlate with clinicopathological features or poor prognosis is of importance in selecting patients who would benefit from radical treatment and to understand the biological characteristics of the disease.
A number of biologic markers that specifically predict the progression and prognosis of pancreatic cancer have been reported, including the K-ras oncogene,10, 11, 12 tumor suppressor DPC4,13, 14, 15, 16 and vascular endothelial growth factor (VEGF).17 Recently, the human tumor-associated antigen RCAS1 (receptor-binding cancer antigen expressed on SiSo cells) has been recognized as another important biological marker expressed on a variety of human cancer cells. This cell membrane molecule was first identified by a mouse monoclonal antibody (22-1-1 antibody) against a human uterine adenocarcinoma cell line, Siso,18 and has been found previously to be a significant prognostic factor not only in cervical cancer but also in non small-cell lung carcinomas.19, 20, 21 Although it has been reported that RCAS1 is also expressed in pancreatic adenocarcinoma cells,22 the prognostic significance of its expression in pancreatic cancer has not been examined to date. As expression of RCAS1 induces apoptotic cell death of receptor-positive immunocytes,23 it has been suggested that upregulation of RCAS1 may play an important role in tumor progression by enabling cancer cells to evade immune surveillance, hence RCAS1 might represent a highly significant prognostic factor distinct from other known factors. The purpose of our study was therefore to perform an immunohistochemical analysis of RCAS1 expression in a cohort of patients with pancreatic adenocarcinoma and to investigate its relationship to clinicopathological variables and prognosis.
PATIENTS AND METHODS
Patients and Tissue Specimens
Surgically resected specimens from 80 patients with pancreatic adenocarcinomas and 7 patients with chronic pancreatitis were studied. All patients underwent surgical resection between 1992 and 1999 in the Second Department of Surgery at the Hokkaido University School of Medicine, the Department of Surgery, Teinekeijinkai Hospital and the Department of Surgery, Hokkaido Gastroenterology Hospital. Among the 80 patients with cancer, 61 patients underwent pancreaticoduodenectomy (Whipple operation), 14 patients underwent distal pancreatectomy and 5 patients underwent total pancreatectomy and all patients received extended radical lymphadenectomy. Of the 23 patients with positive portal vein invasion, 17 patients received portal vein resection. Pancreatic resection was not performed in patients presenting with distant site metastases. Any cases of cystadenocarcinoma or mucin-producing carcinoma were excluded from our study. None of these patients received either radiation or chemotherapy. Pancreatic adenocarcinoma tissues were obtained from 45 men and 35 women with a mean age of 62 years (range 31–83). Median duration of follow-up was 19.8 months (range 3.1–98.0) and 60 patients (75%) died during the follow-up period. All specimens were fixed in 10% formalin and embedded in paraffin wax. Unstained 4-μm sections were then cut from paraffin blocks for immunohistochemical analysis. Histological classification of tumors was based on the World Health Organization criteria. All tumors were staged according to the pTNM pathological classification of the UICC (International Union Against Cancer).24 Seventeen of the tumors were classified as Stage II, 37 as Stage III and 26 as Stage IVA tumors. Thirty of 80 patients had positive resection margin for carcinoma evaluated microscopically (Table I).
Table I. Clinical Characteristics of the Patients with Pancreatic Adenocarcinoma
Number of cases
Depth of invasion
Lymph node metastasis
Tumor diameter (cm)
For immunohistochemical analysis, formalin-fixed and paraffin-embedded specimens were deparaffinized in xylene and dehydrated through a graded series from ethanol to water. Endogenous peroxidase activity was blocked by incubation in 3% hydrogen peroxide solution for 10 min. Specimens were washed in phosphate-buffered saline (PBS pH 7.4) twice and incubated with a protein blocking agent (10% normal goat serum) for 5 min. The primary antibody (anti-RCAS1 mouse monoclonal antibody, Medical & Biological Laboratories Co., Ltd., Nagoya, Japan) was applied at 1:500 dilution and specimens were incubated overnight at 4°C. After 3 additional washes, a biotinylated goat antibody to mouse immunoglobulin (Histofine Simple Stain MAX PO (MULTI); Nichirei Corporation, Tokyo, Japan) was applied at room temperature for 30 min. All reagents were washed in PBS three times and incubated with peroxidase-conjugated streptavidin at room temperature for 30 min. The immunohistochemical reactions were developed in freshly prepared 3,3′-diaminobenzidine tetrahydrochloride (Histofine Simple stain DAB Solution; Nichirei Corporation, Tokyo, Japan) for approximately 15 min and washed in distilled water. Sections were counterstained with hematoxylin for 1 min and mounted in Permount. For negative controls, we used 10% normal mouse serum as the primary antibody.
We selected sections that contained the deepest site in each tumor and counted RCAS1 positive cells with a microscopic field of ×400. Five areas containing highest density of reactive cells were evaluated in each case and the mean percentage of cancer cells with positive staining for RCAS1 was determined. Tissue sections with <5% reactive tumor cells were considered negative and those with 5–25%, 25% to 50% and >50% reactive cells were defined as 1+, 2+ and 3+, respectively. The 25% level was the most significant prognostic effect in overall survival. Therefore, we defined 25% as the cutoff for “low” (−/1+) and “high” (2+/3+) expression. Three investigators of our study independently evaluated the immunohistochemical labeling of all specimens, without knowledge of the patients' background or outcome. Immunoreactivity in each section was represented by the median scores evaluated by 3 investigators.
The correlation of RCAS1 immunoreactivity with patients' clinicopathological variables was analyzed by the chi-square test or Fisher's exact test. The Kaplan-Meier method was used to estimate overall survival. Survival differences according to RCAS1 expression were analyzed by the log-rank test. The influence of variables on survival was assessed using Cox univariate and multivariate regression analysis. The risk ratio and its 95% confidence interval were recorded for each marker. Probability values of less than 0.05 were considered statistically significant in all of the analysis. All analysis was performed with statistical software (StatView version 5.0; SAS Institute Inc., Cary, NC).
Among 80 pancreatic adenocarcinoma specimens, 77 cases (96%) were positive for RCAS1. RCAS1 expression was defined as negative in 3 cases, 1+ in 13 cases, 2+ in 13 cases and 3+ in 51 cases. Patients were classified into 2 groups according to RCAS1 expression. Sixty-four cases (80%) scored 2+ or 3+ were classified as RCAS1 high expression and remainder as low expression. High expression of RCAS1 was found in 11/17 cases (64.7%) of Stage II, 30/37 (81.1%) of Stage III and 23/26 (88.5%) of Stage IVA. RCAS1 was detected both in the cytoplasm and on the plasma membranes of cancer cells, however no reactivity was observed in normal pancreatic cells. (Fig. 1a,b). In all chronic pancreatitis specimens, the ductal cells exhibited no reactivity for RCAS1.
Correlations between RCAS1 expression and various clinicopathological features are summarized in Table II. Borderline correlations between RCAS1 expression were noted for lymph node metastasis and stage (p = 0.0608 and 0.0934, respectively). No significant correlation, however, was found between RCAS1 expression and age, gender, depth of invasion, tumor diameter, surgical margin, lymphatic invasion, venous invasion or histopathological grading.
Table II. Association between RCAS1 Expression and Clinicopathological Variables
Survival curves constructed according to the Kaplan-Meier method are shown in Figures 2 and 3. In 80 patients with pancreatic adenocarcinoma, the survival rates for the patients with high RCAS1 expression were significantly lower than for those with low expression (log-rank test, p = 0.0012; Fig. 2). In selected cases with Stage III and IVA disease (n = 63), the survival rates for patients with high RCAS1 expression were significantly lower than those with low expression (log-rank test, p = 0.0124; Fig. 3).
Univariate Analysis of RCAS1 Expression and Clinicopathologic Variables
Univariate analysis for overall survival using Cox regression analysis identified high RCAS1 expression (p = 0.0023), lymph node metastasis (p = 0.0006), tumor diameter (p = 0.0031) and the surgical margin (p = 0.0019) as significant prognostic predictors. Age, gender, depth of invasion, lymphatic invasion, venous invasion and histopathological grading had no prognostic value (Table III).
Table III. Prognostic Factors in Cox Proportional Hazards Model
Depth of invasion
Lymph node metastasis
Tumor diameter (cm)
2+, 3+/−, 1+
Multivariate Analysis of RCAS1 Expression and Clinicopathologic Variables
Multivariate analysis of the same set of patients was performed for RCAS1 expression and pathological predictors for survival time using the Cox regression model. The results indicated that high RCAS1 expression was an independent unfavorable prognostic factor (risk ratio, 3.090; p = 0.0090). Positive lymph node metastasis, tumor diameter (>3.5 cm) and positive surgical margin also had independent prognostic value, with a risk ratio of 3.074 (p = 0.0012), 1.831 (p = 0.0261) and 2.084 (p = 0.0092), respectively (Table III).
Tumor cells obtain the ability to evade immune surveillance by several strategies that include lack of adequate T cell costimulation,25, 26 downregulation of cell surface MHC class I expression,27, 28, 29 dysfunction of Fas (CD95/APO1)-mediated apoptosis,30, 31, 32 and secretion of immunosuppressive factors, such as transforming growth factor-β (TGF-β).33 In addition to these mechanisms, it has recently been reported that RCAS1 can act as a ligand for a receptor present on activated human T cells and natural killer (NK) cells that inhibits proliferation and induces apoptotic cell death. It was also demonstrated that apoptosis of tumor-infiltrating lymphocytes was induced around RCAS1 positive tumor cells by using the TUNEL method.20
In our study, immunohistochemical analysis revealed that RCAS1 was expressed very frequently, in 96% (77/80 cases) of pancreatic adenocarcinomas. This incidence of RCAS1 expression in pancreatic adenocarcinoma is the highest among reported carcinomas. Previous immunohistochemical studies reported that RCAS1 was expressed in 87.5% of uterine cervical adenocarcinomas, 66% of uterine endometrial adenocarcinomas, 58.8% of ovarian carcinomas and 87.7% of uterine cervical squamous cell carcinomas.18 In non-gynecological cancers, RCAS1 expression was detected in 47.1–74.2% of lung cancers.20, 21 A significant correlation between RCAS1 expression and poor prognosis has also been found in patients with uterine cervical adenocarcinoma19 and with non small-cell lung carcinoma.20, 21
In previous studies RCAS1 expression has been reported to correlate with neoplastic progression. In uterine cervical squamous cell epithelium, RCAS1 was detected in 82.6% of invasive squamous cell carcinomas, 16.7% of microinvasive carcinomas and 20% of carcinoma in situ. RCAS1, however, was not detected in normal uterine cervix and dysplasias.22 In uterine endometrium, RCAS1 was positive in 26% of normal uterine endometrium, 32% of hyperplasias and 68% of uterine endometrial adenocarcinomas.34 In marked contrast to its high incidence in pancreatic adenocarcinoma, we did not detect RCAS1 expression in normal pancreatic duct cells. Thus it seems likely that RCAS1 might also be relevant to neoplastic progression in pancreatic adenocarcinoma. It will worth investigating RCAS1 expression in other types of pancreatic lesions, such as intraductal neoplasias, to clarify the role of RCAS1 in pancreatic cancer progression.35, 36, 37
Based on its proposed immunosuppressive activity, the high incidence of RCAS1 expression might be one of the mechanisms contributing to the dismal outcome of patients with resected pancreatic cancer. Consistent with this hypothesis, we found that the survival rates for pancreatic cancer patients with high RCAS1 expression were significantly lower than for those with low expression (p = 0.0012). The overall 5-year survival rate was 37.9% for patients in the RCAS1 low expression group, compared to 9% for those in the RCAS1 high expression group. Moreover, multivariate analysis demonstrated that high expression of RCAS1 was an independent unfavorable prognostic factor. These findings support the idea that evasion of cancer cells from immune surveillance by expression of RCAS1 may play an important role in determining the course of the patients with pancreatic adenocarcinoma. Therefore, RCAS1 expression in resected specimens may be a useful index of adjuvant therapy for the patients with a high risk of poor prognosis.
Although RCAS1 is a 40-kDa type II membrane protein with an N-terminal transmembrane segment and a coiled-coil structure in the C-terminal portion that is thought to form oligomers, secretion of a soluble form has been confirmed in culture medium of Siso cells (human uterine adenocarcinoma cell line).18, 23 The RCAS1 antigen has also been detected in vaginal discharge from uterine cervical cancer patients.18 This suggests that detection of RCAS1 in the serum or in the pancreatic juice of the patients with pancreatic cancer might also be useful as a highly specific marker for preoperative diagnosis. Moreover, as a number of different tumor-specific antigens have previously been used for gene targeting of virus vectors38, 39 and for selective therapeutic gene expression through use of their cancer cell-specific promoters,40, 41, 42, 43 the high specificity and frequency of RCAS1 expression in patients with pancreatic adenocarcinoma might similarly enable the use of RCAS1 for cancer gene therapy.