Abstract 1387: HER2-positive breast cancer cell lines are particularly sensitive to the novel KDM5 inhibitor GS-701644: Definition of a gene expression model predicting sensitivity to the agent

Abstract

Abstract BACKGROUND Histone methylation controls DNA transcription via modulation of the chromatin structure. Histone methyl-transferases catalyze the transfer of methyl groups to lysine/arginine residues of histone proteins. The demethylation process is controlled by demethylases including members of the KDM family. Alterations in the profiles of histone methylation are observed in the neoplastic cell and pharmacological targeting of this epigenetic process may result in anti-tumor effects. RESULTS GS-701644 is a novel inhibitor of KDM5 histone demethylases. We evaluated the growth inhibitory effects of GS-701644 in 42 breast cancer cell lines. The concentration-dependent effects of GS-701644 (0.001-10 μM) demonstrated that HER2-positivity was a primary determinant of GS-701644 sensitivity. GS-701644 sensitivity was also associated with enrichment in estrogen-receptor-positive (ER+ ) cell lines. Given the sensitivity of HER2-positive (HER2+) cell lines, we evaluated the anti-proliferative action of GS-701644 in combination with the two HER2 targeting drugs, trastuzumab and lapatinib. We exposed HER2+ BT-474 and SK-BR-3 cells to increasing concentrations of GS-701644 and trastuzumab or lapatinib alone or in combination. When BT-474 and SK-BR-3 cells were challenged with trastuzumab or lapatinib alone, a dose-dependent growth inhibition was observed. If the two cells were treated with combinations of GS-701644 and trastuzumab or lapatinib, strong synergistic interactions were demonstrated. In vivo, GS-701644 showed single-agent activity in xenografts of HER2+ BT-474 cells. We determined the basal levels of KDM5A, KDM5B, and KDM5C proteins across our panel of cell lines to look for correlations with sensitivity to GS-701644. No significant association was observed between the amounts of these proteins and sensitivity to GS-701644, promoting a search for other molecular determinants of sensitivity. To achieve this, we analyzed RNA-seq data associated with each breast cancer cell line using a machine learning strategy to define a predictive model of GS-701644 sensitivity. This resulted in the identification of 15 genes whose expression was directly or inversely associated with GS-701644 sensitivity. This gene-expression model predicted responsiveness of cross-validated samples and maintained performance in the test set. Application of the model to the breast cancer cases of public datasets correctly predicted sensitivity to GS-701644 in a large fraction of HER2+ tumors. CONCLUSIONS GS-701644 has potential as a single agent, or in combination with standard-of-care agents, in the treatment of HER2+ breast cancer. The gene-expression model predicting sensitivity to GS-701644 represents a useful diagnostic tool for the selection of patients who may benefit from treatments based on the KDM5 inhibitor. Citation Format: Enrico Garattini. HER2-positive breast cancer cell lines are particularly sensitive to the novel KDM5 inhibitor GS-701644: Definition of a gene expression model predicting sensitivity to the agent [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1387.