Distinct PIK3CA mutation profiles in estrogen receptor (ER) negative breast cancer (BC) subsets.

540 Background: ER negative, progesterone receptor (PR) negative BC is clinically heterogeneous and underlying biology may be complex. We have identified two primary subtypes of ER(-)/PR(-) BC based on genome-wide molecular analysis (Doane et al., 2005, Oncogene, In Press), however the biological mechanisms underlying these molecular differences are not known.,We performed genome-wide expression analysis of 99 primary BC samples and 13 BC cell lines. Total RNA was extracted and hybridized to Affymetrix HG-U133 oligonucleotide microarrays. Unsupervised and supervised analyses were used to investigate gene expression profiles. Predictive models were developed using a supervised k-nearest neighbor technique. Samples were further characterized by immunohistochemistry (IHC) for selected gene products, and HER2 status was determined by IHC and FISH. PIK3CA exons 9 and 20 were sequenced by RT-PCR amplification, direct sequencing, and restriction enzyme digest.,Genes differentially expressed between ER(-)/PR(-) subsets (designated class A and B) included SPDEF, ALCAM, AR, and FGFR4 (p<0.0001). Further study revealed a significant association with class A and the PIK3CA A3140G:H1047R activating mutation among ER(-)/PR(-) BC (p=0.012). HER2 status (gene amplification or 3+ IHC) was not significantly associated with this subtype (p=0.33). MDA-MB-453 corresponded to class A according to molecular profile and PIK3CA mutation status. This cell line demonstrated a proliferative response to androgen (A) in an androgen receptor (AR) dependent and ER-independent manner. In addition the A-induced transcriptional program of MDA-MB-453 significantly overlapped the molecular signature of ER(-)/PR(-) class A human BC (p<0.0001).,Hormonal signaling through the AR may be significantly different in ER(-)/PR(-) molecular subsets. Enhanced AR signaling may be due to cooperation with growth factor signaling through PI3K. The potential molecular crosstalk between activated PI3K and the androgen signaling pathway provides therapeutic opportunities and deserves further study. No significant financial relationships to disclose.