Src pathway activation correlates with treatment resistance in breast cancer and identifies patient subsets predicted to benefit from Src inhibition

PL03 Biomarkers that identify patient tumors with survival pathway activation will be critical for patient selection in clinical trials evaluating the efficacy of novel pathway inhibitors. Furthermore, identifying molecular mechanisms of primary or acquired drug resistance in human cancer will be key to the design of rational combinations of pathway inhibitors with the goal to overcome drug resistance in the clinic. Recent work suggests the oncogenic role of Src involves both dysregulation of integrin signaling and the potentiation of multiple growth factor signaling pathways. Here, we use markers of Src pathway activation (phospho-Src, paxillin, focal adhesion kinase (FAK), p130Cas, phospho-Stat5, and phospho-Stat6) to identify breast cancer patient subpopulations with Src activation in quantitative tissue microarray (TMA) and reverse phase protein array studies. A TMA cohort with 650 patient samples representing the entire breast cancer disease population was analyzed for expression of markers of Src pathway activation as well as Her2, ER and PR. An advantage of automated, quantitative analyses of subcellular antigen localization is that this approach allows for specific assessment of marker activation status in tumors (e.g. membrane localization of phospho-Src versus total Src). In the TMA study, a subset of 57 patients was identified as expressing low levels of ER/PR/Her2 (triple negative breast cancer). In this patient set, subcellular localization of p130Cas and paxillin expression segregate a subgroup of triple negative breast cancer patients with a highly significant decrease in overall survival (Cox chi2, =0.012), suggesting that Src pathway activation may represent a mechanism of treatment resistance in this patient population. A strong direct correlation was observed with Her2 levels of expression and all markers of Src pathway activation (Spearman Rho correaltions for biomarkers range 0.5-0.7) Additional subset analysis indicates that a subset of those patients with either ER or PR expression have evidence of Src pathway activation, and that pathway activation predicted poor prognosis in this group (Cox chi2, p=0.027-0.056). Expression of the Src responsive marker, p130Cas has been associated with endocrine resistance and the expression of p130Cas in the entire patient population was correlated with worse outcome in the subgroup analysis (Cox chi2, p=0.039). To complement these data and build further associations, we generated protein lysates from an independent set of 50 paraffin-embedded, diagnostic breast cancer patient samples and assessed Src pathway marker status by quantitative reverse phase protein arrays using antibodies to Src pathway biomarkers. In these studies, we found similar and robust associations between Src pathway activation and poor prognostic features in a subset of ER/PR positive tumors as well as triple negative breast cancer patients. Together, these pathway biology biomarkers identify similar patient sets and both studies demonstrated the prognostic and potential predictive nature of these markers. These studies together suggest Src pathway activation is a mechanism of treatment resistance in subsets of breast cancer patients, and inhibition of Src signaling in human breast cancer may have a role in overcoming both cytotoxic and endocrine treatment resistance.