P1-06-24: Nuclear Localization of Stat5a Predicts Response to Antiestrogen Therapy and Prognosis of Clinical Breast Cancer Outcome.

Nuclear-localized and tyrosine-phosphorylated Stat5 has been reported as a favorable prognostic marker and predictor of response to antiestrogen therapy in breast cancer. Phospho-Stat5 antibodies do not distinguish between phosphorylated Stat5a and the closely related Stat5b, but Stat5a is considered more critical for normal mammary development than Stat5b. The purpose of this study was to determine whether levels of nuclear-localized Stat5a protein (Nuc-Stat5a) were prognostic of clinical outcome or predictive of antiestrogen response. Stat5a was detected by traditional diaminobenzidine-chromogen immunohistochemistry (IHC) and pathologist scoring or by quantitative immunofluorescence in five archival cohorts of breast cancer. Levels of nuclear-localized Stat5a (Nuc-Stat5a) were evaluated by pathologist scoring of whole tissue sections detected by IHC or automated quantitative analysis (AQUA) of immunofluorescently-labeled tissue microarrays. Levels of Nuc-Stat5a were reduced in invasive breast cancer tissues and lymph node metastases compared to normal tissue and ductal carcinoma in situ when quantified by AQUA (Material I; n=180). Tissues from patients not treated with adjuvant therapy or treated with antiestrogen monotherapy were analyzed according to Nuc-Stat5a status for breast cancer-specific survival (CSS) and time to recurrence (TTR) using univariate and multivariate statistical models, adjusting for clinical features including tumor grade, size, lymph node and ER, PR and Her2 status. In two prognostic cohorts of node-negative breast cancer patients, low expression of Nuc-Stat5a, detected by standard IHC (Material II; n=223) or quantitative analysis (Material III; n=198), was prognostic of poor breast cancer outcome as measured by univariate and multivariate CSS (Material II/III) and TTR (Material II). CSS and TTR analysis of two independent materials of tumors from patients treated with antiestrogen monotherapy and analyzed by standard IHC (Material IV; n=73) or quantitative immunofluorescence (Material V; n=97) indicated that patients whose tumors expressed low levels of Nuc-Stat5a were at a greater than 4-fold risk of antiestrogen therapy failure when adjusted for hormone receptor status and clinical features (multivariate CSS: Material IV HR=4.3 (1.2,15.6), p=0.03; Material V HR=5.0 (1.87,13.06), p=0.001). In conclusion, loss of Nuc-Stat5a is a promising independent marker of poor breast cancer prognosis in node-negative, non-adjuvant treated breast cancer patients. Additionally, Nuc-Stat5a may be a useful clinical tool to predict tumor response to antiestrogen therapy.Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-06-24.

Duke Scholars

Author Terry Hyslop Biostatistics & Bioinformatics, Division of Translational Bi ...