Abstract 3858: Feasibility and efficacy of a precision treatment approach for triple-negative breast cancer in mouse models

Abstract

Abstract 15-20% of human breast tumors are triple negative breast cancer (TNBC), an aggressive and deadly subtype of breast cancer that currently lack targeted therapies, leaving chemotherapy as the only systemic treatment option. There is a pressing need for a better understanding of disease mechanisms of TNBC, and for the development of new treatment options. While activating mutations of PIK3CA are frequently found in ER-positive and Her2-amplified breast cancer, inactivation of lipid phosphatases is more frequent in TNBC. Analysis of large human genomics data in TCGA reveals that heterozygous-loss of INPP4B, a lipid phosphates in the PI3K signaling pathway, is enriched in TNBC subtype, and strongly correlates with loss of ER expression. Guided by human genomics information, we have crossed INPP4B phosphatase deletion mice into TNBC mouse model to determine whether INPP4B-loss cooperates with loss-of-function of p53 and/or Brca1 to promote tumorigenesis in vivo. Our results show dose-dependent increase in tumor development frequency in K14cre; Brca1flox/flox; p53flox/flox mice carrying INPP4B Phosphatase KO allele, HET allele compared to INPP4B WT allele. Importantly, these tumors resemble human TNBC in their pathology, histological patterns and gene expression patterns, providing a valuable platform to test their responsiveness to various therapeutic drugs, including PI3K-inhibitors. Our goal is to generate data that form the basis for a phase I/II clinical study that lead to substantially improved treatment strategies for basal-like breast cancer. Toward this goal, we have banked the endogenous tumors in a way allowing transplantation in nude mice, and have performed randomized drug treatment studies. We found that tumors that are heterozygous or deletion for INPP4B are more sensitive to PI3K-inhibition, suggesting these tumors are more dependent on Pi3K activation to thrive. Although tumors show partial response early on, they inevitably relapsed, begging for an understanding of innate and selected drug-resistant mechanisms. To this end, we have completed large-scale RNAseq and whole exome sequencing (WES), and have performed in-depth analyses on the mouse genomics data to identify genetic alternations including chromosomal number variations (CNVs), mutations and gene fusions. Focusing on gene-fusions, we found that by targeting these fusions, we can achieve better treatment outcomes and in some cases, even complete tumor remission. Significantly TCGA data analyses revealed the presence of similar fusions in human TNBC patients and we are in the process of investigating whether human cells carrying these fusions respond equally well compared to their mouse counterparts. Our approach of precision-medicine guided treatment optimization has lead to substantially improved treatment outcomes. Citation Format: Hui Liu, Charles Murphy, Florian Karreth, Kangkang Yang, Gerburg Wulf, Lewis C. Cantley. Feasibility and efficacy of a precision treatment approach for triple-negative breast cancer in mouse models. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3858.