Abstract PD3-2: Accurate and robust prediction of clinical response to aromatase inhibitors by two weeks of neoadjuvant breast cancer treatment

Background: Aromatase inhibitors (AIs) have an established role in the treatment of estrogen receptor alpha positive (ER+) post-menopausal breast cancer. Response rates are 50-70% in the neoadjuvant setting and lower in advanced disease. There is a need to identify biomarkers to predict response that outperform those currently available, to be able to offer more stratified treatments and improved patient care. Methods: Pre- and on-treatment (at 14 days and 3-months) biopsies were obtained from 89 post-menopausal women with ER+ breast cancer receiving 3 months of neoadjuvant Letrozole. Illumina Beadarray gene expression data (n = 34) were combined with Affymetrix GeneChip data (n = 55) and cross-platform integration approaches developed as part of this study were implemented to combine data. Dynamic clinical response was assessed for each patient using periodic 3D ultrasound measurements performed during treatment. A gene classifier was developed from pre and 14 day gene array expression data to predict response. An independent series from the Royal Marsden was used to validate the classifier. Results: Response to endocrine therapy in the neoadjuvant setting based on the expression of 4 genes has been developed. The classifier comprises baseline expression of an immune signalling gene and an apoptosis related gene, together with 14 day expression of two proliferation genes. Early on-treatment gene changes in combination with pre-treatment gene expression significantly improve predictive power compared to pre-treatment gene expression alone. The classifier had a 96% accuracy in a training dataset (n = 73) and 91% accuracy in an independent validation dataset (n = 44) dataset. Expression of the pre-treatment immune signalling gene alone predicted for response with 85% and 82% accuracy in training and validation datasets respectively. Higher pre-treatment levels of this gene were associated with a significantly better 1 year progression free survival (PFS) (P = 0.0001). In a larger series of patients treated with neoadjuvant Letrozole (n = 129) higher expression of this gene alone was associated with a significantly improved 10 year RFS (p = 0.0359). In a separate tamoxifen treated cohort (n = 212) higher expression of this gene at diagnosis was associated with a significantly improved 5 year (p = 0.0015) and 10 year (p = 0.04) recurrence free survival (RFS). Conclusion: • A 4 gene classifier has been developed and validated to predict response to neoadjuvant Letrozole. • One of the genes identified is a significant predictor in independent data sets of long term RFS in endocrine treated patients. • This new classifier has the potential to predict accurately the benefit of endocrine therapy and has huge potential clinical value. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr PD3-2.