Key points are not available for this paper at this time.
11008 Background: The biological subtypes of breast cancer designated as Luminal A (LumA), Luminal B (LumB), HER2, and Basal have shown prognostic significance. However, risk predictors based on these gene expression profiles have been difficult to clinically implement. Recognizing that there is a continuum in both the spectrum of breast cancer disease and the risk of survival, we sought to develop a supervised risk classier within the context of biological subtypes of breast cancer. Methods: Microarray and real-time quantitative RT-PCR data from 189 samples, procured as fresh frozen and formalin-fixed paraffin-embedded tissues, were used to statistically select prototypes for the biological subtypes of breast cancer. Classification algorithms were developed using 4 independent breast microarray studies consisting of 1,244 cases. A final test set prediction was done using real-time quantitative RT-PCR data generated on a cohort of 279 breast cancers archived for a median of 20.1 years. Results: The biological subtypes predicted on the combined training set showed prognostic significance in all stages of disease (1,244 subjects; p=7.1e-14), node negative disease with no adjuvant systemic therapy (733 subjects; p=6.2e-7), and in patients treated with endocrine therapy (404 subjects; p=1.3e-3). Basal tumors identified using the 50-gene breast classifier had a more complete response to chemotherapy as compared to the same cohort scored as triple negatives. An additional cohort of 279 heterogeneously treated samples was collected from formalin fixed paraffin embedded tissue. Our risk predictor based on distance to subtype centroids provided additional prognostic information and has a linear relationship to prognosis across diverse cohorts. Conclusions: We present an objective method to predict subtypes of breast cancer and generate a continuous risk score based on biological subtypes. The standardized assay can be performed on archived tissue blocks using real-time qRT-PCR, allowing retrospective cohorts and clinical samples to be tested within the framework of specimen processing. Author Disclosure Employment or Leadership Consultant or Advisory Role Stock Ownership Honoraria Research Expert Testimony Other Remuneration Universtiy Genomics
Parker et al. (Tue,) studied this question.