Abstract P5-09-07: Identification of resistance-specific gene expression signatures in a breast cancer patient-derived xenograft with acquired resistance to different endocrine therapies

Abstract

Abstract Background: Acquired resistance to endocrine treatments (ET) occurs in more than 70% of cases of luminal breast cancer (LBC). We used patient derived xenografts (PDX) models of LBC to study molecular changes associated with acquired resistance to different ET modalities. Methods: A PDX model of LBC, established from an early stage BRCA2-mutated breast cancer, was treated with different ET (tamoxifen, fulvestrant, oophorectomy and letrozole) during several months. Tumors escaping to therapies were re-engrafted and maintained under therapy. ET-resistant and parental hormono-responders tumors were analyzed with immunohistochemistry (IHC), RT-PCR and Affymetrix Gene Expression Arrays. Hormono-resistant tumors were additionally studied for their in vivo response to ET, mTOR and PARP inhibitors. Results: From the initially ET sensitive HBCx22 xenograft model (Cottu, BCRT 2012) two resistant models were obtained respectively to tamoxifen (HBCx22-TamR) and to estrogen deprivation (HBCx22-OvaR). Unsupervised clustering of gene expression showed a clear cut separation between parental, TamR and OvaR tumors. Genes differentially expressed in TamR and OvaR tumors compared to parental HBCx22 were only partially overlapping. Up-Regulated genes in both TamR and OvaR tumors (n = 302) were involved in response to wounding, nucleotide metabolism, immune system, adhesion and cell growth. Biological Processes (BP) specifically deregulated in OvaR tumors (n = 380) included embryonic development, antigen presentation, amino acid and lipid metabolism. The top BP specifically regulated in TamR tumors (n = 1059) were response to estrogen and steroid hormones, TGF-b signaling, hypoxia, regulation of cell proliferation, with several strongly up-regulated genes of the histone clusters 1 and 3. Ingenuity Transcription Factor Analysis predicted activation of NFKB, SP1, AP-1 and JUN, and inhibition of ESR1. RT-PCR and IHC analyses confirmed the down regulation of ER controlled genes in the TamR tumors. Expression of ER co-regulators determined by RT-PCR showed that GREB1 was strongly reduced in TamR, while PBX1, GATA3 and FOXA1 were inhibited in OvaR. IHC analysis showed a loss of PTEN expression in HBCx22, with high levels of p-AKT and p-RPS6 in both parental and TamR and OvaR tumors. In vivo ET showed that the TamR xenograft was resistant to all modalities of ET, while OvaR was resistant to estrogen deprivation while retaining some sensitivity to tamoxifen and fulvestrant. Treatment with the mTOR inhibitor RAD001 arrested tumor growth but did not show any additive effect when combined to ET in TamR or OvaR tumors. Conversely, the combination of RAD001 with Olaparib was highly synergistic and induced complete tumor response in 70% of mice. Conclusions: According to the therapeutic selection, tumors derived from a PDX model of ER+ breast cancer show specific resistance patterns and gene expression profiles including disruption in the ER transcriptional program. The analysis of additional resistant tumors established from a second ER+ PDX will be presented at the meeting. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P5-09-07.