Abstract P4-07-02: Expression of Cell Adhesion Molecules Predicts Prognosis in Early Breast Cancer Patients

Purpose: New prognostic and predictive factors are sought for improvement of tailored treatment in early breast cancer. We examined the clinical impact of cell adhesion molecules (CAM): E-cadherin, N-cadherin, Ep-CAM and CEA. Patients and Methods: Our study population (n=574) consisted of all early breast cancer patients primarily treated with surgery in our center between 1985 and 1994. A tissue micro array (TMA) of formalin-fixed paraffin-embedded tumor tissue was immunohistochemically stained for expression of mentioned CAM. The percentage of membranous stained cells was microscopically analyzed. Based on the median score, all CAM were classified in two groups: low expression versus high expression. For CEA, high expression was further subdivided based on the intensity of staining: high expression and highest expression. Results: High expression was seen for E-cadherin, N-cadherin and Ep-CAM in 49%, 46%, 27% of patients respectively. Low expression, high expression and highest expression were found in respectively 48%, 45% and 8% of cases for CEA. Low expression of E-cadherin (p=0.015) and higher expression levels of N-cadherin, Ep-CAM, CEA (p=0.004; 0.046; 0.001 respectively) all resulted in a worse relapse free period (RFP) of patients. Multivariate analysis revealed only E-cadherin and CEA to be independent prognostic variables. A combination variable was created with expression of both markers: (1) E-cadherin high expression, (2) E-cadherin low or CEA low or high expression (3) CEA highest expression. This variable revealed to be an independent prognostic parameter with high discriminative power for RFP (P Conclusion: We have demonstrated that E-cadherin, N-cadherin, Ep-CAM and CEA are of prognostic influence on outcome concerning RFP in breast cancer patients. A combined variable of E-cadherin and CEA expression revealed to have prognostic influence on RFP with high discriminative power and therefore is a candidate parameter for future outcome prediction of patients. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P4-07-02.