Correlation between expression of MUC1 core protein and outcome after surgery in mass-forming intrahepatic cholangiocarcinoma

Authors


Abstract

BACKGROUND

It has been reported that MUC1 is an important prognostic factor in several cancers. This study investigated the importance of MUC1 as a prognostic factor in mass-forming intrahepatic cholangiocarcinoma (m-ICC).

METHODS

In 50 patients with m-ICC who had undergone hepatectomy, expression of MUC1 was investigated. Expression of MUC1 was examined by immunohistochemical staining with monoclonal antibody HMPV, which recognizes the MUC1 core peptide. The immunohistochemical staining patterns of MUC1 were classified into three types: ductal type (the luminal surface membrane of neoplastic cells was stained), cytoplasmic type (the cytoplasm of neoplastic cells was stained dominantly), and negative type.

RESULTS

Expression of MUC1 was detected immunohistochemically in 38 (76%) of 50 cases of m-ICC (ductal type, 18; cytoplasmic type, 20; and negative type, 12). Seventy-five percent of patients with lymph node metastasis had the cytoplasmic type MUC1 expression. Lymph node dissection was performed in only 20 patients, but significant correlation was demonstrated between MUC1 expression and lymph node metastasis (P = 0.0227). The location of MUC1 expression correlated with surgical outcome in m-ICC. Patients with the cytoplasmic type expression showed significantly lower survival rates. Univariate analysis revealed that MUC1 expression was a statistically significant risk factor affecting outcome in m-ICC (P = 0.0028). Furthermore, expression of MUC1 was found to be a statistically significant independent risk factor in multivariate analysis (P = 0.0063).

CONCLUSIONS

The results suggest that evaluation of MUC1 expression may be very useful in predicting the surgical outcome in m-ICC. Cancer 2002;94:1770–6. © 2002 American Cancer Society.

DOI 10.1002/cncr.10398

Intrahepatic cholangiocarcinoma (ICC) is a cancer arising from the epithelium of the “intrahepatic” biliary ducts located on the hepatic side of the first branch of the right and left hepatic duct.1 Intrahepatic cholangiocarcinoma is classified into mass-forming type, periductal infiltrating type, and intraductal growth type by macroscopic appearance.1 These subtypes have different clinical and biologic features and show different surgical outcomes.2–4 Mass-forming type ICC (m-ICC) is a noteworthy tumor similar to hepatocellular carcinoma (HCC), which has a higher curative resection rate by hepatectomy than other forms of ICC.5 However, unlike patients with HCC, the surgical outcome in patients with m-ICC is very poor.6, 7 Therefore, reconsidering the indication of surgery for m-ICC is necessary, and new treatment strategies to improve survival are needed.

MUC1 has attracted attention as a tumor-associated peptide antigen.8–10

MUC1 is a membrane-associated glycoprotein with an extracellular domain consisting of a variable number of highly conserved tandem repeats of 20 amino acids, a transmembrane domain, and a cytoplasmic tail of 69 amino acids.11–15 MUC1 is expressed on the surface of a variety of human normal and neoplastic tissues. MUC1 mucins isolated from tumor and normal cell types differ in their degree of glycosylation. Changes in cell surface glycoproteins during malignant transformation and tumor progression are thought to influence the biologic behavior of cancer cells.16, 17 MUC1 mucin is recognized in the surface membrane of only a few bile duct cells in normal liver. It has been reported that MUC1 is an important prognostic factor in other cancers.18–21 The objective of this study is to evaluate whether MUC1 is an important prognostic factor in m-ICC and to discuss the potentiality of MUC1 as a therapeutic option.

MATERIALS AND METHODS

Patients

Between August 1986 and July 1999, 50 patients, 33 men and 17 women, with m-ICC underwent hepatectomy at our single institute. Patients ranged in age from 30 to 75 years (mean ± standard deviation, 60.3 ± 10.3). They were selected consecutively by reviewing the pathologic findings. Surgical procedures were as follows: 30 patients underwent hepatectomy alone, 12 underwent hepatectomy with extrahepatic bile duct resection, 5 underwent hepatectomy with pancreatoduodenectomy, 10 underwent hepatectomy with partial resection of the portal vein and/or inferior vena cava. The diagnosis of m-ICC was based on histologic features determined in hematoxylin and eosin–stained slides of resected specimens. Patients with combined hepatocellular and cholangiocellular carcinoma were excluded. No patients had certain other diseases such as hepatolithiasis or primary sclerosing cholangitis. M-ICC associated with thorotrast was not included in this study. Before surgery, no patient had undergone chemotherapy or radiotherapy.

Monoclonal Antibody

The expression of MUC1 was examined by monoclonal antibody HMPV(mAb HMPV), which recognizes MUC1 core peptide (mouse immunoglobulin (Ig) G, data sheet from Fujisawa Pharmaceutical Co, Ltd., Tokyo, Japan).

Immunohistochemistry

Immunohistochemical staining was performed on formalin fixed paraffin embedded tissue sections by an immunoperoxidase method using the streptavidin-biotin-peroxidase complex. Each section was deparaffinized with xylene and rehydrated in descending dilutions of ethanol. Endogenous peroxidase activity was blocked by 30 minutes of incubation with 0.3% hydrogen peroxidase in absolute methanol. The sections were incubated with the primary antibody (mAb HMPV, 1:1000 dilution; in PBS) at 4 °C overnight. This was followed by the avidin-biotin procedure with a streptavidin-biotin-peroxidase kit (LSAB; DAKO, Carpinteria, CA). After 12 hours, the sections were incubated for 30 minutes at room temperature in biotinylated secondary antibodies (anti-mouse IgG). The sections then were immersed in the streptavidin-biotin-peroxidase complex for 30 minutes. The reaction products were visualized using 3,3′-diaminobenzidine tetrahydrochloride (DAB; Sigma Chemical Co., St. Louis, MO) and hydrogen peroxide. Finally, the sections were lightly counterstained with hematoxylin.

The results of the antibody staining were classified by the stained parts of neoplastic cells instead of degree of staining as follows: negative type, neoplastic cells were not stained (stained parts were < 5% of whole neoplastic cells); ductal type, the luminal surface membrane of neoplastic cells was stained; cytoplasmic type, both the luminal surface membrane and the cytoplasm of neoplastic cells were stained, but expression in the cytoplasm was dominant (Fig. 1).

Figure 1.

Staining patterns of MUC1 core protein. (A) Ductal type: in the ductal type, the luminal surface membrane of neoplastic cells was stained. (B) Cytoplasmic type: in the cytoplasmic type, both the luminal surface membrane and the cytoplasm of neoplastic cells were stained, but expression in the cytoplasm was dominant.

Statistic Analysis

The relations between clinicopathologic findings, patient survival, and MUC1 expression were studied. Statistical analyses for comparisons between clinicopathologic findings and MUC1 expression were performed using the chi-square test and the Student t test. Patient survival was calculated from the time of surgical resection to either death or most recent contact. The survival curves were generalized by using the Kaplan–Meier method and then were compared by means of the log-rank test. Several clinicopathologic factors were subjected to univariate and multivariate analysis using Cox proportional hazards regression model. Differences were considered significant at P values less than 0.05.

RESULTS

Expression of MUC1 Core Protein

The expression of MUC1 core protein was detected by immunohistochemistry in 38 (76%) of 50 cases of m-ICC (ductal type, 18; cytoplasmic type, 20).

Correlation between MUC1 Expression and Clinicopathologic Features

Correlation between MUC1 expression and clinicopathologic features is summarized in Table 1. Lymph node dissection was performed in only 20 patients, but significant correlation was demonstrated between MUC1 expression and lymph node metastasis (P = 0.0227). Patients with lymph node metastasis had significantly more cytoplasmic type expression.

Table 1. Correlation between Expression Patterns of MUC1 Core Protein and Clinicopathologic Features
FeatureNegative type (%)Ductal type (%)Cytoplasmic type (%)P value
Age (yrs)0.0962
 ≤ 60 (n = 24)8 (33.3)10 (41.7)6 (25.0)
 > 60 (n = 26)4 (15.4)8 (30.8)14 (53.8)
Gender0.7943
 Man (n = 33)7 (21.2)12 (36.4)14 (42.4)
 Woman (n = 17)5 (29.4)6 (35.3)6 (35.3)
Noncancer liver0.4593
 Normal (n = 26)7 (26.9)8 (30.8)11 (42.3)
 Chronic hepatitis (n = 10)3 (30.0)4 (40.0)3 (30.0)
 Cirrhosis (n = 10)1 (10.0)6 (60.0)3 (30.0)
Tumor size (cm)0.0891
 < 4 (n = 11)0 (0)6 (54.5)5 (45.5)
 ≥ 4 (n = 39)12 (30.8)12 (30.8)15 (38.5)
Histologic differentiation0.5318
 Well differentiated (n = 15)2 (13.3)6 (40.0)7 (46.7)
 Moderately differentiated (n = 23)5 (21.7)9 (39.1)9 (39.1)
 Poorly differentiated (n = 12)5 (41.7)3 (25.0)4 (33.3)
Lymph node metastasis0.0227
 Negative (n = 8)3 (37.5)4 (50.0)1 (12.5)
 Positive (n = 12)1 (8.3)2 (16.7)9 (75.0)
Intrahepatic metastasis0.4994
 Negative (n = 30)8 (26.7)12 (40.0)10 (33.3)
 Positive (n = 20)4 (20.0)6 (30.0)10 (50.0)
Vascular invasion0.8808
 Negative (n = 26)7 (26.9)9 (34.6)10 (38.5)
 Positive (n = 24)5 (20.8)9 (37.5)10 (41.7)

Correlation between MUC1 Expression and Cumulative Survival Rate

Median and mean lengths of survival for patients with m-ICC who underwent hepatectomy was 20.0 and 31.9 months, respectively. The 1-, 3-, and 5-year survival rates of patients with m-ICC who underwent hepatectomy were 63.9%, 38.9%, and 29.7%, respectively. The 1-, 3-, and 5-year survival rates for patients with the ductal type, the cytoplasmic type, and the negative type were 87.5%, 65.8%, and 49.4% (median, 47 months); 46.6%, 13.3%, and 6.7% (median, 11 months); and (58.3%, 43.8%, and 43.8% (median, 23 months), respectively. The survival rate of patients with the cytoplasmic type expression was significantly poorer than that of patients with the ductal type expression (P = 0.0046, log-rank test, Fig. 2). However, there was no statistically significant difference between the cytoplasmic type and the negative type.

Figure 2.

Correlation between MUC1 expression and cumulative survival rate (Kaplan–Meier method). The survival rate of patients with the cytoplasmic type MUC1 was poorest and significantly poorer than that of patients with the ductal type MUC1 (log-rank test). However, there were no statistically significant differences between the cytoplasmic type and the negative type.

Univariate Analysis of Prognostic Factors

The univariate analysis of prognostic factors of m-ICC is summarized in Table 2.

Table 2. Univariate Analysis of Prognostic Factors
VariableHazard ratio95% confidence limitsP value
Age (yrs)
 ≤ 60 (n = 24)1
 > 60 (n = 26)1.3360.640–2.7880.4408
Gender
 Man (n = 33)1
 Woman (n = 17)1.7700.753–4.1590.1901
Noncancer liver
 Normal (n = 26)1
 Chronic hepatitis (n = 10)0.4040.117–1.4000.1529
 Cirrhosis (n = 10)1.1250.451–2.8040.8007
Tumor size (cm)
 < 4 (n = 11)1
 ≥ 4 (n = 39)2.8451.069–7.5740.0363
Histologic differentiation
 Well differentiated (n = 15)1
 Moderately differentiated (n = 23)1.9520.746–5.1110.1730
 Poorly differentiated (n = 12)3.9021.363–11.1760.0112
Lymph node metastasis
 Negative (n = 8)1
 Positive (n = 12)2.9960.911–9.8560.0709
Intrahepatic metastasis
 Negative (n = 30)1
 Positive (n = 20)3.9161.684–9.1080.0015
Vascular invasion
 Negative (n = 26)1
 Positive (n = 24)2.0170.956–4.2570.0656
MUC1 core protein
 Cytoplasmic type (n = 20)4.2691.650–11.0420.0028
 Ductal type (n = 18)1
 Negative (n = 12)2.3540.786–7.0490.1259

Univariate analysis revealed that tumor size (P = 0.0363), histologic differentiation (P = 0.0112), intrahepatic metastasis (P = 0.0015), and MUC1 expression (P = 0.0028) were statistically significant risk factors affecting outcome in m-ICC.

Multivariate Analysis of Prognostic Factors

The multivariate analysis of prognostic factors in m-ICC is summarized in Table 3.

Table 3. Multivariate Analysis of Prognostic Factors
VariableHazard ratio95% confidence limitsP value
Tumor size (≥ 4 cm)4.0041.139–14.0710.0305
Histologic differentiation (poorly differentiated)3.8401.116–13.2200.0329
Lymph node metastasis (positive)3.4480.812–14.6450.0935
Intrahepatic metastasis (positive)3.4431.304–9.0920.0126
Vascular invasion (positive)1.4280.615–3.3130.4070
MUC1 core protein (cytoplasmic type)4.3771.517–12.6290.0063

Tumor size (P = 0.0305), histologic differentiation (P = 0.0329), intrahepatic metastasis (P = 0.0126), and MUC1 expression (P = 0.0063) were found to be statistically significant independent risk factors. Furthermore, histologic differentiation (P = 0.0062), intrahepatic metastasis (P = 0.0066), and MUC1 expression (P = 0.0011) were selected as important risk factors by stepwise multivariate analysis (Table 4).

Table 4. Stepwise Multivariate Analysis of Prognostic Factors
VariableHazard ratio95% confidence limitsP value
Histologic differentiation (poorly differentiated)5.40511.6145–18.09570.0062
Intrahepatic metastasis (positive)3.49301.4156–8.61880.0066
MUC1 core protein (cytoplasmic type)4.21241.9309–14.07090.0011

DISCUSSION

In this study, we noticed differences in the location of MUC1 expression in m-ICC.

We studied relations between clinical features, surgical outcomes, and the location of MUC1 expression in m-ICC. The expression of MUC1 has been reported to be associated with outcomes of patients with extrahepatic bile duct carcinoma,22 oral squamous cell carcinoma,23 and breast carcinoma.24 However, reports on the location of MUC1 expression in cancer cells in m-ICC are rare.23, 24 The location of MUC1 expression was classified into the luminal surface membrane (ductal type) and cytoplasm (cytoplasmic type). Surgical outcomes were poorer in patients with the cytoplasmic type MUC1 expression than in those with the ductal type MUC1 expression. The location of MUC1 expression correlated with surgical outcome in m-ICC. Nitta et al. reported that MUC1 mucin antigen was localized in the cytoplasmic vesicles over flat cell surfaces and microvilli in immunogold electron microscopic examination in oral squamous cell carcinoma.23 Ohtani et al. demonstrated that the intracellular localization of the DF3 antigen (MUC1 core protein) in infiltrating breast carcinoma was associated with rough endoplasmic reticulum, the perinuclear space, the Golgi complex, and the cytosol.24 These reports have suggested that the location of MUC1 expression correlates with tumor progression, metastasis, and invasion.23, 24 We speculated that MUC1, which exists at the cell membrane of normal cells, was produced in surplus into the cytoplasm with tumor progression.

Tumor free margins, lymph node metastasis, vascular invasion, mucobilia, and tumor spreading types have been reported as prognostic factors of m-ICC.4, 25 In this study, intrahepatic metastasis and histologic differentiation were evaluated as important prognostic factors along with MUC1 expression. It has been reported that lymph node metastasis is the most important prognostic factor in ICC.26–29 However, the assessment of lymph node metastasis is very difficult in m-ICC. Besides, lymph node dissection may not be performed in some cases of m-ICC because of recognized wide lymph node metastasis, a preoperative diagnosis of HCC, or advanced cirrhosis. In this study, lymph node dissection was performed in only 20 of 50 patients with m-ICC. Therefore, the importance of lymph node metastasis as a prognostic factor was not evaluated.

Hepatectomy remains the only effective treatment for m-ICC, even in patients with advanced-stage tumors. However, no effective postoperative adjuvant therapy affecting survival has been available until recently.30 Recently, immunotherapy was developed as a new option for cancer therapy.31 MUC1 core protein may be useful as a target molecule for immunotherapy. MUC1 targeted cytotoxic T-lymphocyte immunotherapy32 and MUC1-based vaccine immunotherapy33 are expected to provide effective supporting treatment after surgery for patients with m-ICC. Furthermore, MUC1 immunotherapy may become an important treatment strategy against m-ICC as the primary therapy. Because patients with the cytoplasmic type MUC1 expression have an overexpression of MUC1 antigens and poor surgical outcome, they may be good candidates for immunotherapy. When needle biopsy is performed for discriminative diagnosis and judgment of the location of MUC1 expression, evaluation of the MUC1 location may become a reason to select immunotherapy rather than surgery as the primary therapy.

Ancillary