Systematic Identification of Prognostic Biological Pathways in Breast Cancer Molecular Subtypes
Jennifer Kaplan, Stuart Schnitt, Daphne Koller, Andrew H Beck. Beth Israel Deaconess Medical Center, Boston; Stanford University, Palo Alto
Background: Breast cancer is a heterogeneous disease with distinct molecular subtypes including luminal A, luminal B, HER2, and basal-like types. Biological pathways driving breast cancer progression within molecular subtypes are incompletely understood. An improved understanding of subtype-specific pathways will facilitate the identification of new prognostic biomarkers and specific therapeutic targets.
Design: We analyzed eleven publically available breast cancer genome-wide expression profiling datasets with associated recurrence/disease free survival data (total samples = 2123) to identify prognostic biologic pathways within breast cancer molecular subtypes. For each dataset, we assigned each sample to a molecular subtype using the PAM50 subtype predictor. We then rank-normalized the expression value of each gene across patients of each molecular subtype for each data set. We pooled the rank-normalized expression values for patients of a subtype across the eleven data sets to create four new meta-data sets containing rank-normalized values for patients from one of the four molecular subtypes. These values were used to compute every gene's univariate association with survival separately within the four subtypes. After performing univariate survival analyses, we performed gene set enrichment analysis (GSEA) separately on the four lists of gene survival statistics to identify biologic pathways significantly associated with prognosis in the molecular subtypes.
Results: Proliferation-associated pathways were most strongly associated with poor survival in ER+ breast cancer (in both luminal A and luminal B subtypes). Conversely, proliferation-associated pathways were associated with improved prognosis in the molecular HER2 type. Extracellular-matrix remodeling-associated pathways were significantly associated with poor prognosis in ER- breast cancer (both HER2 and basal-like groups). In luminal B, HER2, and basal-like subtypes, inflammation-related pathways were associated with improved survival (all FWER p < 0.05).
Conclusions: The results of this study show that diverse biological pathways are associated with prognosis in breast cancer molecular subtypes. These results provide new insights into the pathways that drive tumor progression within molecular breast cancer subtypes and this, in turn, may lead to the development of novel subtype-specific prognostic markers and therapeutic targets. Future studies will evaluate the relationship of these subtype-specific pathways with chemotherapy response.
Wednesday, March 21, 2012 1:00 PM
Poster Session VI # 14, Wednesday Afternoon