Automated Quantification and Clinical Implications of Src, Ezrin, and Tks5 in Breast Cancer

Abstract

Breast cancer (BC) is one of the leading causes of cancer-related deaths in Canadian women. Aggressive BCs (e.g. triple-negative subtype; TN) present a clinical challenge as defined biomarkers, particularly those indicative of unique cancer-associated signaling pathways, are needed to improve prognostication and prediction of therapeutic response. Overexpression of Src and its substrates, Ezrin and Tks5, have been associated with poor prognosis in many cancers. We have previously shown that Ezrin regulates proteolytic-independent invasion, while others have shown that Tks5 is associated with proteolytic-dependent invasion. Thus, expression of Ezrin versus Tks5 in BC cases may represent different invasion modalities. Additionally, immunofluorescence (IF)-based technologies may provide a more quantitative and objective approach for analysis of biomarker expression and subcellular compartmentalization compared to immunohistochemistry (IHC). In this study, I hypothesize that expression and subcellular localization of Src, Ezrin and Tks5, have improved prognostic significance in BCs, compared to current clinico-pathological parameters. To assess this, I optimized an IF-based automated quantification analysis (AQUA) system to measure subcellular expression in a 63-patient BC cohort and tested associations with clinico-pathological data. This thesis presents that: 1) Expression of Src and Ezrin increased, but that of Tks5 decreased in breast tumours compared to normal breast. 2) Src and Ezrin localized to the apical regions of normal breast epithelia but shifted to the cytoplasm in breast tumours. Tks5 exhibited a granular basal expression in normal breast epithelia, and is weakly expressed in tumour cellular compartments. 3) In our 63-patient cohort, Src and Ezrin had significant correlations with multiple clinico-pathological parameters, including TN status and lymphovascular invasion. 4) Clinico-pathological associations with IF-based AQUA scoring are directly comparable to conventional manual IHC scoring. Our study supports the role of Src and Ezrin as potential prognostic biomarkers for BC.