Automated Quantitative Analysis of E-Cadherin Expression in Lymph Node Metastases Is Predictive of Survival in Invasive Ductal Breast Cancer

Open Access
Publisher Website
Google Scholar
Abstract
Purpose: The tumor suppressor adhesion molecule E-cadherin is believed to have an anti-invasive role in breast cancer. Lymph node involvement is the best prognostic marker known, yet there is variability in outcome among node-positive patients. We investigated the relationship between E-cadherin expression in primary invasive ductal tumors and corresponding nodal metastases, and determined the prognostic value of E-cadherin expression in node-positive breast cancer. Experimental Design: Membrane E-cadherin expression was studied by immunohistochemical staining of tissue microarrays with fluorescent-labeled antibodies. An objective method of automated quantitative analysis (AQUA) was used. AQUA uses cytokeratin to define pixels as breast cancer (tumor mask) within the array spot, and measures E-cadherin expression using a Cy5-conjugated antibody within the mask. Results: We employed a tissue microarray containing 207 primary and matched nodal metastases suitable for AQUA analysis. There was no significant difference in mean staining intensity between the primary and nodal specimens (P = 0.8). A scattergram was generated which identified a subset of patients (25%) with high E-cadherin expression in nodal metastases, and this top quartile had improved survival (P = 0.028). On univariate analysis, increased E-cadherin expression in nodal metastases was strongly associated with improved survival (P = 0.007), whereas expression in primary tumors was not (P = 0.13). On multivariate analysis, nodal E-cadherin expression retained its independent association with survival, as did tumor size and HER2/neu status. Conclusions: Strong E-cadherin expression in lymph node metastases was highly predictive of improved survival. This suggests that expression of adhesion molecules at metastatic sites portends less aggressive tumor behavior.

This publication has 29 references indexed in Scilit: