Characterization of triple-negative breast cancer utilizing microarray-based gene-expression profiling

Abstract

e22014 Background: Triple-negative breast cancer (TNBC) is characterized by lack of expression of estrogen receptor (ER), progesterone receptor (PR) and HER-2. This makes it one of the most challenging subgroups for clinical decision-making. Methods: We analyzed gene-expression profiles of 200 node-negative breast cancer patients utilizing the Affymetrix HG-U133A array. After performing an unsupervised hierarchical cluster analysis using 2579 genes selected for variable expression within our dataset, we constructed metagenes for five dominant cluster (basal-like, luminal, proliferation, T-cell, B-cell). The interrelation of the median expression of these metagenes between TNBC and those cancer specimens expressing at least one of those receptors was analysed with the Wilcoxon-Mann-Whitney test. An intrinsic gene list was used to define basal-like breast cancer (BLBC). Sensitivity and specificity of TNBC to correctly classify BLBC was calculated. Metastasis-free survival (MFS) at 5 years was calculated according to Kaplan Meier. Results: Of the 200 node-negative breast cancer patients, 33 (16.5%) were triple-negative and 20 (10%) basal-like. Sensitivity of TNBC to correctly predict BLBC was 100%, specificity was 93.1%. TNBC showed an overexpression of the basal-like metagene (P=7.905e-13), an inverse relation with the luminal metagene (p=1.151e-12) and had higher proliferation and higher expression of T-cell (P=6.316e-5) and B-cell (P=3.551e-6) metagene. A higher expression of the B-cell metagene was associated with longer MFS in both TNBC (P=0.048) and BLBC (P=0.041). Conclusions: Characterization of TNBC reveals marked differences in gene-expression. Overexpression of mRNA transcripts related to the humoral immune system might serve as a protective factor in this particular subgroup. No significant financial relationships to disclose.