Search

Purpose and methods In human basal-like breast cancer, mutations and deletions in TP53 and BRCA1 are frequent oncogenic events. Thus, we interbred mice expressing the CRE-recombinase with mice harboring loxP sites at TP53 and BRCA1 (K14-Cre; p53f/f Brca1f/f) to test the hypothesis that tissue-specific deletion of TP53 and BRCA1 would give rise to tumors reflective of human basal-like breast cancer. Results In support of our hypothesis, these transgenic mice developed tumors that express basal-like cytokeratins and demonstrated intrinsic gene expression features similar to human basal-like tumors. Array comparative genomic hybridization revealed a striking conservation of copy number alterations between the K14-Cre; p53f/f Brca1f/f mouse model and human basal-like breast cancer. Conserved events included MYC amplification, KRAS amplification, and RB1 loss. Microarray analysis demonstrated that these DNA copy number events also led to corresponding changes in signatures of pathway activation including high proliferation due to RB1 loss. K14-Cre; p53f/f Brca1f/f also matched human basal-like breast cancer for a propensity to have immune cell infiltrates. Given the long latency of K14-Cre; p53f/f Brca1f/f tumors (~ 250 days), we created tumor syngeneic transplant lines, as well as in vitro cell lines, which were tested for sensitivity to carboplatin and paclitaxel. These therapies invoked acute regression, extended overall survival, and resulted in gene expression signatures of an anti-tumor immune response. Conclusion These findings demonstrate that this model is a valuable preclinical resource for the study of human basal-like breast cancer. Electronic supplementary material The online version of this article (10.1007/s10549-018-5061-y) contains supplementary material, which is available to authorized users.

The PI3K/Akt signaling axis contributes to the dysregulation of many dominant features in breast cancer including cell proliferation, survival, metabolism, motility, and genomic instability. While multiple studies have demonstrated that basal-like or triple-negative breast tumors have uniformly high PI3K/Akt activity, genomic alterations that mediate dysregulation of this pathway in this subset of highly aggressive breast tumors remain to be determined. In this study, we present an integrated genomic analysis based on the use of a PI3K gene expression signature as a framework to analyze orthogonal genomic data from human breast tumors, including RNA expression, DNA copy number alterations, and protein expression. In combination with data from a genome-wide RNA-mediated interference screen in human breast cancer cell lines, we identified essential genetic drivers of PI3K/Akt signaling. Our in silico analyses identified SOX4 amplification as a novel modulator of PI3K/Akt signaling in breast cancers and in vitro studies confirmed its role in regulating Akt phosphorylation. Taken together, these data establish a role for SOX4-mediated PI3K/Akt signaling in breast cancer and suggest that SOX4 may represent a novel therapeutic target and/or biomarker for current PI3K family therapies.

Background: Cyclin dependent kinases (CDK) comprise a family of proteins which are activated upon cyclin binding to promote cell cycle progression. Historically, CDK family inhibitors have had limited utility in the clinic due to 1) toxicities associated with broad spectrum CDK inhibition; and 2) varied and unpredictable efficacy due to lack of a patient selection strategy. The recent approval of three selective CDK4/6 drugs (palbociclib, ribociclib and abemaciclib) has revived the pursuit of therapies against a related interphase kinase, CDK2. Cyclin E gene (CCNE) alterations are prevalent in cancers with high unmet medical need and are emerging as a potential mechanism of clinical resistance to targeted therapies (e.g. CDK4/6 therapies). Methods: Here we report preclinical validation leading to the development of a selective CDK2 inhibitor for the treatment of cancers harboring CCNE alterations. We highlight advanced, orally bioavailable compounds exhibiting single-digit low nanomolar biochemical and cellular potency (pRB), exquisite CDK family selectivity (CDK1, 4, 6, 7 and 9), and whole kinome selectivity. We have characterized the phenotypic and mechanistic consequences of targeting CDK2, employing both genetic and pharmacologic methods across cell lines and xenograft models (e.g. OVCAR-3, MKN-1, HCC1569). Results: In a panel of cell lines, CCNE1-amplified lines exhibited profound sensitivity to selective CDK2 inhibition by halting cells at the G1/S phase of the cell cycle. Upon further inspection, CDK2 inhibition induced multiple markers of senescence, e.g. SA-b-gal and IL-6, in CCNE1-amplified but not in CCNE non-amplified cell lines. Consistent with this, these compounds exhibit robust anti-tumor activity in multiple CCNE1-amplified xenograft models, which was sustained after removal of treatment with these compounds. Conclusions: These data provide a strong rationale for advancing this class of compounds toward clinical development in CCNE-altered cancers. Citation Format: Yoon J. Choi, Steve Wenglowsky, Victoria Brown, Neil Bifulco, Yeon S. Choi, Jian Guo, Megan Hatlen, Joseph Kim, Tim LaBranche, Riadh Lobbardi, Emanuele Perola, Emily Rozsahegyi, Michelle Maynard, Phil Ramsden, Grace Silva, Faith Stevison, Richard Vargas, Ruduan Wang, Doug Wilson, Rich Woessner, Dean Zhang, Rob Meissner, Klaus Hoeflich, Marion Dorsch. Development of a selective CDK2-E inhibitor in CCNE driven cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1279.

Hematopoietic progenitor kinase 1 (HPK1, MAP4K1) is a serine/threonine kinase that has been demonstrated to have suppressive effects across a range of immune cells, including T cells and dendritic cells. Loss of MAP4K1 kinase activity is sufficient to enhance T cell receptor (TCR) signaling resulting in robust anti-tumor immunity alone and in combination with checkpoint inhibition. These data support that MAP4K1 is a novel and attractive target for cancer immunotherapy. We have designed a series of potent, selective, and orally bioavailable inhibitors of MAP4K1. Treatment of primary human T cells or peripheral blood with either BLU2069 or BLU6348 was able to inhibit phosphorylation of pSLP76, a scaffolding protein that regulates MAPK downstream of the TCR. In addition, we show that compound treatment can enhance cytokine secretion and proliferation in human T cells in response to TCR crosslinking. The therapeutic benefit of MAP4K1 inhibition alone and in combination with anti-PD-L1 was evaluated in multiple syngeneic mouse tumor models including MCA205, MC38 and EMT-6. Treatment with either compound alone led to a reduction in tumor growth that was further enhanced when combined with anti-PD-L1 therapy. When tumors were grown in immunocompromised mice (MCA-205) or in the setting of CD8+ T cell depletion (MC-38), the anti-tumor effect of BLU2069 and BLU6348 respectively was lost, confirming the importance of immune cells in compound mediated antitumor effects. We further show that MCA205 tumors harvested from mice treated with BLU2069 had increased intratumoral CD8+ T cell infiltration, resulting in enhanced CD8/Treg ratios. In addition, transcriptional analysis of tumor lysates showed that BLU2069 significantly increased genes associated with an effector phenotype. These data support that pharmacological inhibition of MAP4K1 reduced tumor burden and enhanced antitumor immunity in preclinical tumor models. Finally, we show that MAP4K1 inhibition can enhance CD3/CD28-induced IL2 and IFNγ in human tumor infiltrating lymphocytes (TILs) generated from melanoma or non-small cell lung cancer (NSCLC) primary tumors. This work describes the identification of potent small molecule inhibitors of MAP4K1 which could be novel therapeutic agents and induce an effective immune response either alone or in combination with approved checkpoint inhibitors. Citation Format: Kerrie Faia, Alberto Toso, Kristina Fetalvero, Marly Roche, Steven Bench, Erin O'Hearn, Qiongfang Cao, Kerry-Ann Bright, Debora Paduraru, Andrea Romagnani, Weifan Weng, Tina Zimmermann, Michael Burke, Joshua Close, Luke Green, Joseph Kim, Chandra Miduturu, Alison Ribeiro, Marina Bacac, Sylvia Herter, Emanuele Perola, Michael Sheets, Jan Eckmann, Gordon Heidkamp, Tary Traore, Erik Gerson, Rich Woessner, Carsten Wolter, Felix Scheuplein, Nisha Perez, Timothy LaBranche, Grace Silva, Chaoyang Ye, Caitlin Utt, Stefan Gross, James R. Bischoff, Marion Dorsch, Tim Guzi, Klaus Hoeflich, Jason Brubaker. MAP4K1 inhibition enhances immune cell activation and anti-tumor immunity in preclinical tumor models [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1717.

Hepatocellular carcinoma (HCC) is a prevalent human cancer with rising incidence worldwide. Human HCC is frequently associated with chronic liver inflammation and cirrhosis, pathophysiological processes that are a consequence of chronic viral infection, disturbances in metabolism, or exposure to chemical toxicants. To better characterize the pathogenesis of HCC, we used a human disease‐relevant mouse model of fibrosis‐associated hepatocarcinogenesis. In this model, marked liver tumor response caused by the promutagenic chemical N‐nitrosodiethylamine in the presence of liver fibrosis was associated with epigenetic events indicative of genomic instability. Therefore, we hypothesized that DNA copy number alterations (CNAs), a feature of genomic instability and a common characteristic of cancer, are concordant between human HCC and mouse models of fibrosis‐associated hepatocarcinogenesis. We evaluated DNA CNAs and changes in gene expression in the mouse liver (normal, tumor, and nontumor fibrotic tissues). Additionally, we compared our findings to DNA CNAs in human HCC cases (tumor and nontumor cirrhotic/fibrotic tissues) using publicly available data from The Cancer Genome Atlas (TCGA). We observed that while fibrotic liver tissue is largely devoid of DNA CNAs, highly frequently occurring DNA CNAs are found in mouse tumors, which is indicative of a profound increase in chromosomal instability in HCC. The cross‐species gene‐level comparison of CNAs identified shared regions of CNAs between human fibrosis‐ and cirrhosis‐associated liver tumors and mouse fibrosis‐associated HCC. Our results suggest that CNAs most commonly arise in neoplastic tissue rather than in fibrotic or cirrhotic liver, and demonstrate the utility of this mouse model in replicating the molecular features of human HCC.

Introduction: Fibrolamellar Carcinoma (FLC) is a rare primary liver malignancy, affecting children and young adults without chronic liver disease. FLC tumors are largely resistant to chemotherapy, making the identification of effective treatment options urgently needed. Recent genomic data strongly suggest that DNAJB1-PRKACA kinase fusions are the drivers of the vast majority of FLC cases. However, it has not been assessed whether FLC tumors remain dependent on DNAJB1-PRKACA expression and whether PRKACA inhibition could be a therapeutic approach for FLC. Here we summarize the preclinical evaluation of PRKACA as a potential therapeutic target for FLC. Methods: We established a xenograft model from a FLC- patient and then developed inducible PRKACA shRNA cell lines from this model. We also designed potent tool compounds that selectively inhibit the PRKACA protein to assess PRKACA as a potential therapeutic target for FLC. Results: We characterized a patient-derived xenograft (PDX) model of FLC (LI5132) and confirmed DNAJB1-PRKACA fusion expression and constitutive PRKACA pathway activation measured by phospho-VASP. The model also shows fibrolamellar type histology by H&E staining and expression of typical FLC markers like cytokeratin 7 and CD68 by IHC. Using inducible PRKACA-specific shRNA cell lines from this PDX model we demonstrated that the FLC transcriptional gene signature correlates strongly with expression of the DNAJB1-PRKACAfusion protein. Importantly, we demonstrated for three inducible PRKACA shRNA cell-line-derived xenograft models that the in vivo tumor growth remained dependent on DNAJB1-PRKACA fusion expression (TGI-72%-78%, day 22). PRKACA knockdown tumors displayed reduced Ki67 index (6.4 %) when compared to non-induced controls (37.1 %) further confirming that proliferation of the tumors depends on the fusion expression. To investigate the PRKACA catalytic dependency of the FLC model, we designed potent and selective PRKACA inhibitors based on starting points from our proprietary kinase inhibitor library. These investigational compounds are the first selective and potent PRKACA inhibitors and provide excellent tools to assess in vitro and in vivo PRKACA dependency. These compounds achieved potent PRKACA pathway inhibition and dose-dependent inhibition of FLC-specific gene expression, including genes such as carbamoyl phosphate synthetase (CPS1) and forkhead box C1 (FoxC1). We established a pharmacokinetic/ pharmacodynamic relationship and demonstrated in vivo PRKACA pathway inhibition in PDX tumors, as measured by phospho-VASP. Importantly, oral delivery of a potent and selective PRKACA inhibitor achieved up to 80% PRKACA kinase inhibition and led to statistically significant FLC tumor growth inhibition (54%, day 34) on a tolerated schedule. These data demonstrate that FLC depends on PRKACA kinase activity. Conclusion: This study is the first evaluation of PRKACA kinase inhibition as a therapeutic approach for FLC. The results from these preclinical experiments provide strong evidence that FLC depends on PRKACA catalytic activity and that novel PRKACA inhibitors may significantly decrease tumor growth in vivo. Citation Format: Stefanie S Schalm, Erin O’Hearn, Kevin Wilson, Timothy LaBranche, Grace Silva, Lucian DiPietro, Neil Bifulco, Richard Woessner, Nicolas Stransky, Darshan Sappal, Adam Shutes, Robert Campbell, Riadh Lobbardi, Michael Palmer, Joseph Kim, Stephen Miller, Marion Dorsch, Christoph Lengauer, Timothy Guzi, Vivek Kadambi, Andrew Garner, Klaus P Hoeflich. Evaluating PRKACA as a therapeutic target for Fibrolamellar Carcinoma [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr B127. doi:10.1158/1535-7163.TARG-19-B127

Fibrolamellar carcinoma (FLC) is a unique liver cancer primarily affecting young adults and characterized by a fusion event between DNAJB1 and PRKACA. By analyzing RNA-sequencing data from The Cancer Genome Atlas (TCGA) for >9,100 tumors across ~30 cancer types, we show that the DNAJB1-PRKACA fusion is specific to FLCs. We demonstrate that FLC tumors (n = 6) exhibit distinct messenger RNA (mRNA) and long intergenic non-coding RNA (lincRNA) profiles compared to hepatocellular carcinoma (n = 263) and cholangiocarcinoma (n = 36), the two most common liver cancers. We also identify a set of mRNAs (n = 16) and lincRNAs (n = 4), including LINC00473, that distinguish FLC from ~25 other liver and non-liver cancer types. We confirm this unique FLC signature by analysis of two independent FLC cohorts (n = 20 and 34). Lastly, we validate the overexpression of one specific gene in the FLC signature, carbonic anhydrase XII (CA12), at the protein level by western blot and immunohistochemistry. Both the mRNA and lincRNA signatures support a major role for protein kinase A (PKA) signaling in shaping the FLC gene expression landscape, and present novel candidate FLC oncogenes that merit further investigation.

Mutations in PIK3CA, the gene encoding the p110α catalytic subunit of phosphoinositide 3-kinase (PI3K) have been shown to transform human mammary epithelial cells (MECs). These mutations are present in all breast cancer subtypes, including basal-like breast cancer (BLBC). Using liquid chromatography-tandem mass spectrometry (LC-MS/MS), we identified 72 protein expression changes in human basal-like MECs with knock-in E545K or H1047R PIK3CA mutations versus isogenic MECs with wild-type PIK3CA. Several of these were secreted proteins, cell surface receptors or ECM interacting molecules and were required for growth of PIK3CA mutant cells as well as adjacent cells with wild-type PIK3CA. The proteins identified by MS were enriched among human BLBC cell lines and pointed to a PI3K-dependent amphiregulin/EGFR/ERK signaling axis that is activated in BLBC. Proteins induced by PIK3CA mutations correlated with EGFR signaling and reduced relapse-free survival in BLBC. Treatment with EGFR inhibitors reduced growth of PIK3CA mutant BLBC cell lines and murine mammary tumors driven by a PIK3CA mutant transgene, all together suggesting that PIK3CA mutations promote tumor growth in part by inducing protein changes that activate EGFR.

Controversy exists regarding the impact of CYP2D6 genotype on tamoxifen responsiveness. We examined loss of heterozygosity (LOH) at the CYP2D6 locus and determined its impact on genotyping error when tumor tissue is used as a DNA source.Genomic tumor data from the adjuvant and metastatic settings (The Cancer Genome Atlas [TCGA] and Foundation Medicine [FM]) were analyzed to characterize the impact of CYP2D6 copy number alterations (CNAs) and LOH on Hardy Weinberg equilibrium (HWE). Additionally, we analyzed CYP2D6 *4 genotype from formalin-fixed paraffin-embedded (FFPE) tumor blocks containing nonmalignant tissue and buccal (germline) samples from patients on the North Central Cancer Treatment Group (NCCTG) 89-30-52 tamoxifen trial. All statistical tests were two-sided.In TCGA samples (n =627), the CYP2D6 LOH rate was similar in estrogen receptor (ER)-positive (41.2%) and ER-negative (35.2%) but lower in HER2-positive tumors (15.1%) (P.001). In FM ER+ samples (n = 290), similar LOH rates were observed (40.8%). In 190 NCCTG samples, the agreement between CYP2D6 genotypes derived from FFPE tumors and FFPE tumors containing nonmalignant tissue was moderate (weighted Kappa = 0.74; 95% CI = 0.63 to 0.84). Comparing CYP2D6 genotypes derived from buccal cells to FFPE tumor DNA, CYP2D6*4 genotype was discordant in six of 31(19.4%). In contrast, there was no disagreement between CYP2D6 genotypes derived from buccal cells with FFPE tumors containing nonmalignant tissue.LOH at the CYP2D6 locus is common in breast cancer, resulting in potential misclassification of germline CYP2D6 genotypes. Tumor DNA should not be used to determine germline CYP2D6 genotype without sensitive techniques to detect low frequency alleles and quality control procedures appropriate for somatic DNA.

Elucidating the molecular drivers of human breast cancers requires a strategy capable of integrating multiple forms of data and an ability to interpret the functional consequences of a given genetic aberration. Here we present an integrated genomic strategy based on the use of gene expression signatures of oncogenic pathway activity (n=52) as a framework to analyze DNA copy number alterations in combination with data from a genome-wide RNAi screen. We identify specific DNA amplifications, and importantly, essential genes within these amplicons representing key genetic drivers, including known and novel regulators of oncogenesis. The genes identified include eight that are essential for cell proliferation (FGD5, METTL6, CPT1A, DTX3, MRPS23, EIF2S2, EIF6 and SLC2A10) and are uniquely amplified in patients with highly proliferative luminal breast tumors, a clinical subset of patients for which few therapeutic options are effective. Our results demonstrate that this general strategy has the potential to identify putative therapeutic targets within amplicons through an integrated use of genetic, genomic, and genome-wide RNAi data sets.

The Cancer Genome Atlas (TCGA) Research Network has profiled and analyzed large numbers of human tumors to discover molecular aberrations at the DNA, RNA, protein and epigenetic levels. The resulting rich data provide a major opportunity to develop an integrated picture of commonalities, differences and emergent themes across tumor lineages. The Pan-Cancer initiative compares the first 12 tumor types profiled by TCGA. Analysis of the molecular aberrations and their functional roles across tumor types will teach us how to extend therapies effective in one cancer type to others with a similar genomic profile., National Cancer Institute (U.S.), National Human Genome Research Institute (U.S.)

Background: The molecular characterization of primary breast cancers has led to signatures identifying risk of future metastasis and survival; however the underlying biology driving metastasis development is largely unknown. Methods: Utilizing a Rapid Autopsy Program, we have collected 61 metastatic breast cancer tumors from 7 individuals (4 triple negative, 2 HER2+, 1 ER+/HER2-) including primary tumors and 3-6 metastases/patient. We performed mRNA and DNA exome sequencing. We next used DawnRank, a novel network-based method that integrates DNA and RNA data to identify computationally determined "driver" genes (i.e. a DNA variant that significantly alters its gene expression-network) in each individual sample. Phylogenetic tree and clonal analysis were also performed, with the computationally determined drivers mapped onto these trees. Results: The breast cancer primaries were molecularly subtyped as 5 Basal-like, 1 HER2-Enriched, and 1 Luminal A; in all cases, the metastases clustered immediately adjacent to their primary tumor by hierarchical clustering analysis. Widespread DNA copy number alterations identified in the primary tumors were typically maintained throughout metastasis. On average, 1.9 ± 1.3% of DNA copy number altered genes, and 2.4 ± 0.95% of the somatic mutations per tumor were identified as "drivers" by DawnRank. There were an average of 199 ± 72 total drivers per tumor due to copy number alterations (amplifications or deletions) and 12 ± 23 drivers per tumor from somatic mutations. Phylogenetic tree analysis demonstrated that the majority of DNA copy number events occurred early in tumor development. Founding clones were defined as genetic events present in the primary and all matched metastases. Chr5q13 loss and TP53 mutation were the only consistent alterations in the founding clones of all 7 patients. Drivers on chr5q13 identified in this cohort include CCNB1, CDK7, and TAF9. Among the basal-like patients, all 5 patients' TP53 mutations were identified as a driver by DawnRank. 39% and 20% of drivers from copy number gains and losses, respectively, were identified in the primary tumor, while another 34% and 30% were not seen in the primary but were present in more than 1 metastasis within each patient. Metastasis-enriched copy number drivers not seen in any primary included FLT1, MAP2KR, and ARNT. 38% of the drivers resulting from somatic mutations were established in the primary and maintained in metastases. Of the remaining drivers from somatic mutation, only 18% were shared among metastases but not seen in the primary while 47% were not seen in any other tumor within a given patient (i.e. private to a single sample). TP53, PSEN1, CDC27, HDAC1, and BRCA1 were somatic mutation drivers established early in metastatic development, while CCNH was a consistent late driver. Conclusions: We present a novel computationally determined genetic "driver" analysis of matched breast cancer primaries and multi-organ metastases. In this cohort, our results suggest that most genetic drivers in a single tumor are based on copy number aberrations, are established early, and are maintained in metastases. In contrast to copy number, drivers from somatic mutations are acquired later, and most of the metastases continued to acquire new genetic driving features. Citation Format: Siegel MB, He X, Chen M, Hou JP, Garrett AL, Dye JB, Silva GO, Usary JE, Moylan VJ, Brady CM, Ma J, Thorne LB, Hoadley KA, Parker JS, Anders CK, Carey LA, Perou CM. Identification of early versus late drivers of breast tumors and metastasis. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr S4-01.

Nature 523, 313–317 (2015); doi:10.1038/nature14583 In this Article, author Aurelien A. Serandour should have been listed with one middle initial only. This has been corrected in the online versions.

Background: Genomic characterization of human breast tumors has resulted in the identification of at least five distinct tumor subtypes (Luminal A and B, HER2-enriched, Claudin-low, and Basal-like). These subtypes each show their own unique biology, which correlates with distinct patient outcomes. Basal-like Breast Cancer (BLBC) is associated with high grade, BRCA1 mutation status, younger patient age, and increased frequency in African Americans. In order to identify novel therapeutic strategies for BLBC, we developed a genetically engineered mouse model based upon two common human BLBC genetic events, namely BRCA1 and TP53 loss. Results: We developed a genetically engineered mouse (GEM) model of breast cancer by inactivating TP53 and BRCA1 genes using Cre-LOX technology in the mammary gland (K14-cre / TP53 L/L / BRCA1 L/L). As part of the characterization of this GEM model, we performed gene expression profiling and DNA copy number analysis using aCGH arrays on the Basal-like tumors from this model (25/27 total tumors were Basal-like). We found a number of chromosomal regions that were similarly altered and found to be specific to mouse and human BLBC including gains of mouse chromosome 1q and 12q and losses in chromosomes 3F1 and 15q. Subsequently two tumors from the murine Basal-like group were taken and expanded into an orthotopic syngeneic model where they passaged in wild type FVB mice. These two tumors lines were then tested extensively with chemotherapeutics (Doxorubicin, Carboplatin, Cytoxan, Paclitaxel) and biologically targeted agents (MEK inhibitor/AZD6244, PI3K Inhibitor/BEZ235, PARP Inhibitor/ABT888 and mTOR Inhibitor/Everolimus). The DNA damaging agents exhibited exceptional efficacy (i.e. ABT888, Cytoxan, and Carboplatin), whereas the biologically targeted agents tended to at best, only slow tumor growth, even in combination. Conclusions: Genetic engineering of mice using TP53 and BRCA1 loss results in tumors of predominantly the Basal-like subtype phenotype. As expected, these tumors were exquisitely sensitive to DNA damaging chemotherapeutics including cytoxan and carboplatin, yet generally insensitive to the 3 tested kinase inhibitors. Novel combinations of targeted agents and chemotherapeutics are now being tested. Citation Format: Cristina M. Contreras, Grace O. Silva, Adam D. Pfefferle, Charles M. Perou. Targeted agents and chemotherapeutic treatments of genetically engineered Basal-like breast cancer mouse model. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2307. doi:10.1158/1538-7445.AM2015-2307

Introduction: Nearly half of metastatic triple negative breast cancer (TNBC) patients develop brain metastases (BM) and face a poor prognosis. The blood-brain barrier (BBB) prevents many treatments from reaching intracranial tumors, and there are no FDA-approved systemic therapies to treat TNBC BM. In this study, we evaluated the efficacy of BBB-permeable, clinically-available inhibitors of MEK and identified rational co-target pathways in preclinical models of intracranial (IC) TNBC. Methods: In vitro IC50s, synergy, and siRNA screens (700 kinase genes) were conducted in 4 human-derived TNBC lines (SUM149, MDA-MB-468, MDA-MB-436, MDA-MB-231Br). We evaluated the efficacy of the MEK1/2 inhibitor AZD6244 (AZD), the pan-PI3K inhibitor BKM120 (BKM), and the PDGFR inhibitor Pazopanib (Pazo) in IC TNBC mouse models. Tumor burden was monitored via bioluminescence, and IC tumors were frozen for gene expression analyses using custom human 4×44K Agilent microarrays or of kinome activity profiles using multiplex kinase inhibitor beads and mass spectrometry. To determine drivers of AZD sensitivity, DNA copy number data (Broad CCLE) was analyzed using SWITCHplus to identify copy number alterations that differ between sensitive (n = 8) vs. resistant (n = 12) TNBC lines based on their IC50s (Sanger Cancerxgene). Results: In vitro, SUM149 and 231Br TNBC cells exhibited lower (40 uM). Several genes synthetically enhanced lethality in SUM149 and 231Br cells: PI3K genes and PDGFRα/β with AZD, and MAPK/MAP2K/MAP3K genes with BKM, suggesting MEK+PI3K and MEK+PDGFR inhibition as rational combinations. AZD plus BKM or Pazo were synergistic in vitro in sensitive cell lines. In vivo, AZD reduced tumor burden and improved survival in the SUM149 (72 vs. 45 days in controls, p < 0.005) and 231Br (37 vs. 30 days, p < 0.02) models, with no benefit in the other two models. Single agent BKM or Pazo resulted in little to no improvement in any model. However, in AZD-sensitive models, combined AZD+BKM inhibition increased survival (SUM149: 87 vs 45 days, p < 0.0001; MDA-MB-231Br: 52 vs 30 days, p < 0.001) as did AZD+Pazo (SUM149: 88 vs. 35 days, p < 0.0001). Several DNA segments were significantly altered in sensitive vs. resistant TNBC cell lines. Notably, MEK-pathway genes were lost in the resistant lines. Ongoing work will complete characterization of therapies in all models in vitro and in vivo and will compare genetic, transcriptional, and kinome activity alterations. Conclusions: TNBC models exhibit different innate sensitivities to the BBB-permeable MEK inhibitor AZD6244. In sensitive models, AZD improves survival and reduces intracranial tumor burden, and rational combined inhibition of PI3K or PDGFR further increases survival. Identification of predictive biomarkers will enable translation of our results to biomarker-driven clinical trials for patients with TNBC BM. Citation Format: Amanda E.D. Van Swearingen, Marni B. Siegel, Maria J. Sambade, Shivani Sud, Samantha M. Miller, Grace Silva, Ryan E. Bash, Charlene M. Santos, David B. Darr, Brian Golitz, Joel S. Parker, C. Ryan Miller, Gary L. Johnson, Carey K. Anders. Combination therapy with MEK inhibition is efficacious in intracranial triple negative breast cancer models. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2579. doi:10.1158/1538-7445.AM2015-2579

Background: Breast cancer is not a single disease, but rather is a collection of unique diseases with distinct molecular mechanisms and clinical characteristics. Although substantial progress has been made in the development and directed administration of new cancer therapies, the reality is that most patients with advanced tumors will succumb to their disease. Since recent studies have suggested that the activity of altered oncogenic pathways may be an important determinant of the biology of the tumor and potential response to therapy, identifying mechanisms driving key oncogenic pathways is paramount to understanding the transformation process and ultimately enabling the development of therapeutic regimens that can match the complexity of a given patient's tumor. Methods: A series of gene expression signatures (n = 52) that measure oncogenic pathway activity in human breast tumors were utilized as a conceptual framework to integrate genome-wide level expression and DNA copy number data in order to identify genetic drivers of oncogenic pathway activity. Results: A panel of pathway-based gene expression signatures was applied to a collection of nearly 500 breast tumor samples from the TCGA project. In addition to expression data, each tumor had matched DNA copy number data that were used to identify chromosomal alterations associated with each pathway signature. This strategy was validated by identifying associations between pathway activity and copy number alterations (CNA) of core pathway components including the correlation between ERBB2, cMYC, and E2F1 amplification and the Her2, Myc, and E2F1 signatures, respectively. These results were further confirmed by analyzing Reverse Phase Protein Array (RPPA) data to identify functional activation of the pathway as illustrated by changes in protein expression including increased phosphorylated (p) AKT, Myc and pEGFR relative to the PI3K, Myc, and EGFR signatures, respectively. Finally, to identify those amplified genes that are essential for cell viability in a pathway-dependent manner, we analyzed a dataset of breast cancer cell lines with available mRNA expression and genome-wide shRNA proliferation data. By integrating DNA copy number and shRNA analyses, we identified SOX4 as a novel regulator of PI3K signaling in basal-like breast cancers. We determined that SOX4 is amplified in approximately 40% of basal-like breast tumors with high PI3K signaling, which is comparable to the frequency of PIK3CA or PTEN alterations. By analyzing RPPA data, we demonstrate that tumors with SOX4 amplification have increased expression of core PI3K protein components including pAKT, thus further confirming the association between SOX4 amplification and aberrant PI3K/AKT signaling. Conclusions: In this study, we used a panel of pathway-based gene expression signatures as a means to integrate four disparate forms of genomic data to identify drivers of oncogenic pathway activity, including SOX4 amplification as a novel modulator of PI3K signaling in basal-like breast cancers. This novel strategy is able to identify oncogenic events in a context specific manner, which is essential for understanding mechanisms driving tumorigenesis and for the development of personalized therapeutic strategies. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr S4-01.

A large number of DNA copy number alterations (CNAs) exist in human breast cancers, and thus characterizing the most frequent CNAs is key to advancing therapeutics because it is likely that these regions contain breast tumor ‘drivers’ (i.e., cancer causal genes). This study aims to characterize the genomic landscape of breast cancer CNAs and identify potential subtype-specific drivers using a large set of human breast tumors and genetically engineered mouse (GEM) mammary tumors. Using a novel method called SWITCHplus, we identified subtype-specific DNA CNAs occurring at a 15 % or greater frequency, which excluded many well-known breast cancer-related drivers such as amplification of ERBB2, and deletions of TP53 and RB1. A comparison of CNAs between mouse and human breast tumors identified regions with shared subtype-specific CNAs. Additional criteria that included gene expression-to-copy number correlation, a DawnRank network analysis, and RNA interference functional studies highlighted candidate driver genes that fulfilled these multiple criteria. Numerous regions of shared CNAs were observed between human breast tumors and GEM mammary tumor models that shared similar gene expression features. Specifically, we identified chromosome 1q21-23 as a Basal-like subtype-enriched region with multiple potential driver genes including PI4KB, SHC1, and NCSTN. This step-wise computational approach based on a cross-species comparison is applicable to any tumor type for which sufficient human and model system DNA copy number data exist, and in this instance, highlights that a single region of amplification may in fact harbor multiple driver genes. Electronic supplementary material The online version of this article (doi:10.1007/s10549-015-3476-2) contains supplementary material, which is available to authorized users.

We analysed primary breast cancers by genomic DNA copy number arrays, DNA methylation, exome sequencing, messenger RNA arrays, microRNA sequencing and reverse-phase protein arrays. Our ability to integrate information across platforms provided key insights into previously defined gene expression subtypes and demonstrated the existence of four main breast cancer classes when combining data from five platforms, each of which shows significant molecular heterogeneity. Somatic mutations in only three genes (TP53, PIK3CA and GATA3) occurred at >10% incidence across all breast cancers; however, there were numerous subtype-associated and novel gene mutations including the enrichment of specific mutations in GATA3, PIK3CA and MAP3K1 with the luminal A subtype. We identified two novel protein-expression-defined subgroups, possibly produced by stromal/microenvironmental elements, and integrated analyses identified specific signalling pathways dominant in each molecular subtype including a HER2/phosphorylated HER2/EGFR/phosphorylated EGFR signature within the HER2-enriched expression subtype. Comparison of basal-like breast tumours with high-grade serous ovarian tumours showed many molecular commonalities, indicating a related aetiology and similar therapeutic opportunities. The biological finding of the four main breast cancer subtypes caused by different subsets of genetic and epigenetic abnormalities raises the hypothesis that much of the clinically observable plasticity and heterogeneity occurs within, and not across, these major biological subtypes of breast cancer.