Overexpression of epithelial cell adhesion molecule (Ep-CAM) is an independent prognostic marker for reduced survival of patients with epithelial ovarian cancer

Abstract

Objective.

Currently available clinical and molecular factors provide still an insufficient prognostic and predictive assessment for patients with epithelial ovarian cancer (EOC). To identify a potential molecular target and prognostic/predictive factor for EOC, we investigated in a retrospective study the prognostic value of Ep-CAM overexpression in EOC.

Methods.

We assessed by immunohistochemistry the expression of the Ep-CAM antigen on tissue microarrays containing paraffin-embedded tissue samples of 199 patients with documented EOC. Patients were operated for ovarian cancer in the period between June 1980 and January 2000.

Results.

We observed a rate of Ep-CAM overexpression of 68.8%. Ep-CAM overexpression was significantly related to a decreased overall survival (P = 0.036). The prognostic power of Ep-CAM overexpression was particularly strong in patients with stage III and IV disease. In fact, in this subgroup, median overall survival was twofold higher in patients without as compared to patients with Ep-CAM overexpression (46 vs. 23 months, P < 0.01). Univariate analysis revealed a correlation with histologic grade. We observed a significantly higher rate of Ep-CAM overexpression (83.5%) in grade 3 tumors. Histologic subtypes associated with a higher rate of Ep-CAM overexpression were serous carcinoma, squamous cell carcinoma, undifferentiated carcinoma, clear cell carcinoma, and endometrioid carcinoma. Cox regression analysis showed Ep-CAM overexpression to be an independent prognostic marker (P = 0.037, RR = 1.64).

Conclusions.

This retrospective analysis demonstrates for the first time an independent prognostic value of Ep-CAM overexpression in patients with EOC. Ovarian cancer patients with Ep-CAM overexpressing tumors are frequent and would qualify for treatment with Ep-CAM-specific immunotherapeutic approaches.

Introduction

Ovarian cancer is a cancer type, which frequently relapses after treatment with conventional procedures. The majority of patients with epithelial ovarian cancer (EOC) are diagnosed with advanced disease involving upper abdomen, pleural space, and para-aortic lymph nodes (for review, see Cannistra et al. [1]). Despite the highly lethal nature of EOC, the clinical course of advanced disease can be difficult to predict in an individual patient. Prognostic predictors are currently based on clinical and histopathological features including: FIGO stage, histologic grade of tumor, histologic subtype, age at diagnosis, and residual tumor after surgical treatment [2].

Ep-CAM (17-1A, GA733-2, KSA, KS1/4, 323/A3, CD326) is a calcium-independent homophilic cell adhesion molecule of 39–42 kDa which is expressed by the majority of epithelial tissues [3], [4]. The cDNA for Ep-CAM was cloned in 1989 and called KSA [5]. The Ep-CAM antigen is unique in that it is not a member of any of the major families of adhesion molecules such as cadherins, selectins, or integrins. It is a type I transmembrane glycoprotein consisting of an extracellular domain with two EGF-like repeats and a short cytoplasmic domain of 26 amino acids (for review see Balzar et al. [6]). The gene was mapped to human chromosome 2 [7], [8], [9].

Ep-CAM is abundantly and homogeneously expressed on human carcinomas of different origins [10], [11]. Recent immunohistochemical studies of prostate cancer and cervical intraepithelial neoplasia have shown that Ep-CAM expression can increase with disease progression and proliferation [12], [13]. This apparent overexpression has also been described in 35–42% of patients with invasive breast cancer and was a strong predictor of poor disease-free and overall survival [14], [15], [16]. Similar correlations between Ep-CAM overexpression and disease progression could be observed in patients suffering from gallbladder carcinoma [17]. In line with these results, silencing Ep-CAM gene expression with Ep-CAM short interfering RNA (siRNA) resulted in decreased cell proliferation of human breast cancer cell lines [18]. Furthermore, Ep-CAM signaling seems to impact cell proliferation through upregulation of the proto-oncogene c-myc [19]. In fact, overexpressed Ep-CAM has high oncogenic potential allowing cells to grow in soft agar and independent of serum growth factors. Beside its function as adhesion molecule and potential oncogene, Ep-CAM can inhibit CD4⁺ T-cell-dependent immune responses and thus enable tumor cells to evade T-cell-mediated anti-tumor immunity [20].

The Ep-CAM antigen has attracted major interest as a target for passive and active immunotherapy. Anti-tumor responses have been observed in metastatic colorectal cancer after treatment with the Ep-CAM specific monoclonal antibody edrecolomab [21]. Adjuvant treatment of radically resected Dukes' C colorectal cancer patients using the same antibody induced a 32% relative reduction in mortality as compared to the results of surgery alone [22]. However, as compared to chemotherapy, treatment with the murine antibody edrecolomab in a phase III study was shown to be less effective [23]. A fully human anti-Ep-CAM antibody called adecatumumab (MT201) with higher antibody-dependent cellular cytotoxicity (ADCC), longer half-life and reduced immunogenicity compared to edrecolomab [24], [25] is currently tested in phase II clinical trials in breast and prostate cancer patients. Many other approaches are using Ep-CAM as target antigen including immunotoxins and vaccines.

Despite these promising results, only few data on Ep-CAM expression pattern in epithelial ovarian cancer patients are available. One group described a significant higher Ep-CAM expression in ovarian cancer as compared to normal ovarian tissue [26]. These results are consisted with the finding that ovarian cancer contains more Ep-CAM mRNA as compared to normal ovarian tissue [27]. Intriguingly, ovarian cancer patients presented higher amounts of natural anti-Ep-CAM antibodies in the serum as compared to healthy donors. The aim of this study was to evaluate the rate of Ep-CAM overexpression, as defined by our previously described criteria [14] and to test its potential prognostic value in patients with EOC.