Abstract
Abstract Basal-like breast cancer is a molecular subtype of breast cancer generally thought to have a universally poor prognosis. Subsequent studies examining the long-term outcome in thousands of patients with basal-like breast cancer have shown that these patients can be separated into two clinically distinct groups: those likely to experience a systemic recurrence and succumb to their disease within the first 5 years and those expected to show excellent long term survival. The ability to distinguish between these two sub-groups (good and poor prognosis) of basal-like breast cancer patients at the time of initial diagnosis would permit tailoring more aggressive therapeutic regimens to those patients with an inherently poorer prognosis and conversely to avoid such therapy in patients with a more indolent course. We aimed to identify a gene signature that could predict the clinical outcome of basal-like breast cancer patients. To this end we mined publicly available human breast tumor gene expression profiling data and identified patients with basal-like breast cancer. We divided these patients into training and validation sets to identify and confirm the accuracy of a prognostic signature. We identified 137 basal-like breast tumors among 995 breast tumor gene expression profiles. We used 85 of these samples as a training group and identified an optimal 14-gene signature, which accurately identified patients that experienced poor and good long-term survival. We confirmed the accuracy of our gene signature on a 49 patient independent validation set. Importantly, we also confirmed the capacity of our signature to predict outcome in a chemotherapy naïve 27 patient sub-set of the 49 patients validation set. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3663. doi:1538-7445.AM2012-3663
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have