Abstract P4-05-04: Proteomic patterns unravel a new luminal-A breast cancer molecular subgroup with prognostic value

Abstract

Abstract Background: Breast cancer is a heterogeneous disease including a variety of entities with different genetic background. Estrogen receptor-positive, HER2 negative tumors (ER+) usually have a favorable outcome, although some patients eventually relapse, which suggests some heterogeneity within this category. In the last years, proteomic approaches have been incorporated to the study of clinical samples as a way to complement the information provided by gene analysis. Shotgun proteomics allows measuring over 1,000 proteins in clinical samples. In the present study, we combined genomic and proteomic techniques to characterize a set of breast tumors. Methods: The study population consisted of 102 patients with lymph-node positive breast cancer who had received anthracycline-based adjuvant chemotherapy. Protein extracts from FFPE samples were prepared in 2% SDS buffer and digested with trypsin. SDS was removed from digested lysates, and resulting peptides were analyzed in an Orbitrap Velos. Protein abundance was calculated on the basis of the normalized spectral protein intensity (LFQ intensity) using MaxQuant. A prognostic protein signature was built. Findings were verified using whole genome gene expression data from 1,141 patients included in public repositories. To this purpose, the protein signature was converted to a gene signature. Data analysis was done using MeV, BRBArray Tools, R and Cytoscape software suites and Uniprot (http://www.uniprot.org/) and DAVID (http://david.abcc.ncifcrf.gov) webtools. Results: We identified 3,000 protein groups in FFPE breast cancer samples and selected 1,000 that were identified at least in 75% of the samples. Significance Analysis for Microarrays analysis revealed 224 protein groups differentially expressed between ER+ and triple-negative (TN) samples (False Discovery Rate set at <0.001). Hierarchical clustering analyses of protein expression showed that some ER+ samples had a protein expression profile similar to that of TN samples: patients with TN-like tumors had a clinical outcome similar to those with TN disease. Gene ontology analyses unraveled a reduced expression of proteins related with cellular adhesion in the TN-like and the TN groups. A TN-like predictive protein signature was built, converted to a gene signature and evaluated in the whole-genome expression data. The signature had prognostic value in patients with luminal-A breast cancer. This prognostic information was independent from that provided by standard genomic tests for breast cancer, such as MammaPrint, OncoType Dx and the 8-gene Score. Conclusions: Proteomic profiling showed that cellular adhesion is a differential process between ER+ and TN breast cancer, and is reduced in the TN tumors. A group of ER+ breast tumors with reduced cellular adhesion was identified (TN-like). Patients with this luminal-A, TN-like breast cancer type had a poor outcome. This prognostic information was complementary to that offered by genomic tests such as OncoType. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P4-05-04.