Research Project 1: Development of PI3K Inhibitors as Single Agents or in Combination with MEK Inhibitors for Breast Cancer

研究项目1的重点是确定反应预测和耐药机制， 磷脂酰肌醇3-激酶（PI 3 K）抑制剂，通过使用广泛选择的PDX试验进行 激素受体阳性（HR+）/HER 2-和三阴性乳腺癌（TNBC）的乳腺癌模型 亚型PI 3 K抑制剂是用于乳腺癌的有吸引力的治疗剂，这是因为PI 3 K抑制剂在乳腺癌中的频繁发生。 激素受体阳性（HR+）中的PIK 3CA突变和TNBC中的PTEN缺失。然而，单药PI 3 K 抑制剂仅产生适度的抗肿瘤活性，并且pan-PI 3 K抑制剂的功效通常受到以下因素的限制： 临床试验中的剂量限制性毒性。因此，亚型特异性PI 3 K抑制剂具有极大的兴趣，以最大化 靶向抑制，耐受性提高。开发PI 3 K抑制剂的主要挑战是 确定可能从这些药物中获益最多的患者以及亚型特异性抑制剂 是合适的。临床试验集中在有限的标志物，包括PIK 3CA突变和PTEN状态， 并不能一致地预测反应。需要进行深入的生物标志物研究，但通常很困难， 由于研究材料的数量和质量以及无法获得连续肿瘤， 病人的活检。相比之下，患者来源的异种移植物（PDX）模型提供了几乎无限的肿瘤生长。 用于生物标志物开发的治疗前和治疗后深入基因蛋白质组学分析的资源， 确定合理药物组合的耐药机制。我们假设乳腺癌 泛和亚型特异性PI 3 K抑制剂的PDX试验，将辨别个体肿瘤的亚型依赖性， 获得预测性生物标志物。我们将使用copanlisib，一种有效的泛PI 3 K抑制剂，具有抗 主要是PI 3 K α与PI 3 K β，以及AZD 8186，一种特异性PI 3 K β抑制剂，两种药物均在NCI-IND中 portfolio.此外，根据文献证据和我们来自buparlisib PDX试验的初步数据，我们 假设添加MEK抑制剂司美替尼（也是NCI IND药物）将改善对 copanlisib和/或AZD 8186的协同抗肿瘤作用 当在我们的乳腺癌PDX模型中将pan-PI 3 K抑制剂与MEK抑制剂组合时， Aim 1将进行PDX试验。100个乳腺癌PDX模型，由ER+ HER 2-和TNBC亚型组成 将基于传代次数、基因组稳定性和肿瘤大小来选择从PDX核心中可获得的 特色为了富集目标人群，至少30个具有PIK 3CA突变的PDX模型和至少 本PDX试验将纳入30个具有PTEN无效表型的PDX模型。每种PDX模型将传代 12只免疫缺陷小鼠，并随机分配到6个治疗组（每个治疗组n=2）， 媒介物、copanlisib、AZD 8186、司美替尼、copanlisib +司美替尼或AZD 8186+司美替尼。几 肿瘤反应标准将用于对药物治疗敏感与耐药的分类。这些包括 最近发表的mRECIST标准和传统的肿瘤生长抑制%（包括溶媒） 治疗组目标2将确定候选和新的基因蛋白质组学预测反应的每一个单一的代理 或通过分析全局基因组数据（全外显子组和RNA Seq）和蛋白质组数据进行联合治疗 通过多激酶抑制剂珠质谱法（MIB-MS）生成，用于无偏发现候选物 生物标志物。该分析将包括重点确定离群，差异表达的生物标志物之间 敏感和耐药肿瘤。目的3将对治疗后样本进行激酶组分析（MIB-MS），以评估 药物诱导的信号传导变化，以辨别研究治疗的作用机制，并确定 内在的和治疗诱导的适应性生存机制。 该项目将得到WU-PDTC建立的数据收集基础设施的支持，包括 药物治疗、小鼠耐受性和肿瘤体积随时间变化。此外，生物信息学能力 将利用WU-PDTC提供的试验解释和基因蛋白质组学分析， 对PI 3 K抑制剂的反应和耐药机制的预测因子。此外，项目1的团队成员 我将定期与项目2的团队成员进行信息交流，以改进研究方法 成果共享。本研究的长期目标是建立一个标准化的PDX试验平台， 用于临床测试的有前景的药物和生物标志物对。

Research project 1 focuses on identifying response predictors and resistance mechanisms for phosphatidylinositide 3-kinase (PI3K) inhibitors through the conduct of a PDX trial using a broad selection of breast cancer models of hormone receptor positive (HR+)/HER2- and triple negative breast cancer (TNBC) subtypes. PI3K inhibitors are attractive therapeutic agents for breast cancer because of the frequent occurrence of PIK3CA mutations in hormone receptor positive (HR+) and PTEN loss in TNBC. However single agent PI3K inhibitors have yielded only modest anti-tumor activity and the efficacy of pan-PI3K inhibitors are often limited by dose limiting toxicities in clinical trials. Isoform-specific PI3K inhibitors are therefore of great interest to maximize target inhibition with improved tolerability. A major challenge in the development of PI3K inhibitors have been the identification of patients who may benefit the most from these agents and for whom isoform specific inhibitors are appropriate. Clinical trials have focused on limited markers including PIK3CA mutation and PTEN status, which do not consistently predict response. In-depth biomarker research is needed, but is often difficult in the clinical trial setting due to the quantity and quality of the study material and the inability to obtain serial tumor biopsies from patients. In contrast, patient-derived xenograft (PDX) models provide nearly unlimited tumor resources for pre- and post-treatment in-depth genoproteomic analysis for biomarker development and determination of resistance mechanisms for rational drug combinations. We hypothesize that a breast cancer PDX trial of pan- and isoform-specific PI3K inhibitors, will discern isoform dependency of individual tumors to derive predictive biomarkers. We will use copanlisib, a potent pan-PI3K inhibitor with activity against predominantly PI3Kα versus PI3Kβ, and AZD8186, a specific PI3Kβ inhibitor, both agents are in the NCI-IND portfolio. In addition, based on literature evidence and our preliminary data from a PDX trial of buparlisib, we hypothesize that adding the MEK inhibitor selumetinib, also an NCI IND agent, will improve response to copanlisib and/or AZD8186 based on literature evidence and the synergistic anti-tumor effects we observed when combining a pan-PI3K inhibitor with a MEK inhibitor in our PDX models of breast cancer. Aim 1 will conduct the PDX trial. 100 breast cancer PDX models comprised of ER+HER2- and TNBC subtypes available from the PDX core will be selected based on passage number, genomic stability and tumor characteristics. To enrich for the target population, at least 30 PDX models with PIK3CA mutation and at least 30 PDX models with PTEN null phenotype will be included in this PDX trial. Each PDX model will be passaged to 12 immune-deficient mice and randomly assigned to 6 treatment groups (n=2 each treatment group) to receive either vehicle, copanlisib, AZD8186, selumatinib, copanlisib + selumetinib or AZD8186 + selumetinib. Several tumor response criteria will be used for classification of sensitive vs resistant to drug(s) therapy. These include the recently published mRECIST criteria and the traditional % tumor growth inhibition which includes a vehicle treated group. Aim 2 will identify candidate and novel genoproteomic predictors of response for each single agent or combination therapy by analyzing global genomic data (whole exome and RNA Seq) and proteomic data generated by multi-kinase inhibitor bead mass spectrometry (MIB-MS) for unbiased discovery of candidate biomarkers. The analysis will include a focus on identifying outlying, differentially expressed biomarkers between sensitive and resistant tumors. Aim 3 will perform kinome profiling (MIB-MS) of post treatment samples to assess drug-induced signaling changes in order to discern mechanisms of action of the study treatments and to identify intrinsic and treatment-induced adaptive survival mechanisms. This project will be supported by the infrastructure established by the WU-PDTC for data collection, including drug treatment, mice tolerability and tumor volumes changes over time. In addition, the bioinformatics capability provided by the WU-PDTC for trial interpretation and genoproteomic analysis will be leveraged to derive predictors of response and resistance mechanisms to PI3K inhibitors. In addition, team members of Project 1 will interact regularly with team members of Project 2 for information exchange to improve research approach and results sharing. The long term goal of this research is to set up a standardized PDX trial platform to identify promising drug(s) and biomarker pairs for clinical testing.