Abstract Aromatase inhibitors (AI) have evolved over the last decade into an effective therapeutic regime for postmenopausal women with primary or advanced breast cancer. Despite their remarkable success in the clinic, intrinsic resistance to therapy occurs in a proportion of patients, while other patients who respond initially to treatment will relapse with recurrent disease. Previous studies suggest that this may be due, at least in part to estrogen receptor (ER) hypersensitivity. We undertook ER ChIPseq analysis on AI resistant cell model and data for this suggests that ER transcriptional regulation alone may not be responsible for the development of the resistant phenotype. We examined the role of the established ER coactivator protein AIB1. AIB1 has previously been associated with initiation of breast cancer and resistance to endocrine therapy. In tamoxifen treated patients, expression of AIB1 in conjunction with an activated HER2 cascade has been associated with treatment resistance and early disease recurrence. By contrast, we have observed that AIB1 alone can predict response to AIs. In our TMA the expression of AIB1 associated with disease recurrence (p = 0.025) and reduced disease free survival time (p = 0.0471) in patients treated with AIs as first-line therapy. Reflecting increased growth factor activity reported in AI resistance, AIB1 expression associated with the growth factor second messenger signaling proteins, p-Src and pERK1/2, but not the receptor HER2. These results suggest that AIB1 may utilize additional transcription factors other than ER to drive endocrine resistance. Additionally, we show that AIB1 is highly expressed in AI resistant metastases; therefore, monitoring AIB1 expression may be useful to screen for disease progression and detect disease advancement before metastases appear. Our studies of cell line models of AI resistance suggest that AIB1 may play a functional role in aggressive, migratory, phenotype of AI resistance. We have generated cell line models of resistance to letrozole (LetR) and anastrozole (AnaR). Our resistance models have higher levels of AIB1 and have increased migratory capacity. Interestingly, knockdown of AIB1 reduces the migratory capacity of the resistant cells. Furthermore, we have observed that AIB1 regulation of ER target genes is selectively enhanced in AI resistant cells in a promoter specific context. AIB1 recruitment to ER target genes such as pS2 and Myc becomes insensitive to letrozole. By contrast, AIB1 recruitment to cyclinD1 retained letrozole sensitivity. Our evidence suggests that steroidal regulation of transcription factors such as Jun and Fos may contribute to this promoter-specific regulation of ER target genes. We establish a role for AIB1 in AI-resistant breast cancer and describe a new mechanism of ERalpha/AIB1 gene regulation which could contribute to the development of an aggressive tumour phenotype. We provide evidence of a central role for AIB1 in regulating selective ER transcriptional activity and driving tumour recurrence in AI treated patients. Tackling the emerging problem of AI resistance in a timely fashion will enable us to tailor existing therapies and improve outcome in specific patient groups before disease recurrence becomes a clinical issue. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P6-04-21.