INTEGRATING OMICS AND QUANTITATIVE IMAGING DATA IN CO-CLINICAL TRIALS TO PREDICT TREATMENT RESPONSE IN TRIPLE NEGATIVE BREAST CANCER

在临床联合试验中整合组学和定量成像数据来预测三阴性乳腺癌的治疗反应

• 批准号: 10241425
• 负责人: Michael T. Lewis
• 金额: $ 63.29万
• 依托单位: BAYLOR COLLEGE OF MEDICINE
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-19 至 2024-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10241425
• 关键词: Address; Algorithms; Animals; Biological; Biological Markers; Biopsy; Blood Vessels; Breast Cancer Patient; Breast Cancer Treatment; Cancer Biology; Cancer Patient; Carboplatin; Caring; Cellularity; Clinical; Clinical Trials; Communities; Consensus; Coupled; Data; Development; Disease; Disease Progression; Enrollment; Evaluation; Funding; Genomics; Goals; Healthcare; Human; Image; Imaging technology; Immunotherapy; In complete remission; Informatics; Machine Learning; Magnetic Resonance Imaging; Malignant Neoplasms; Mass Spectrum Analysis; Methodology; Methods; Modeling; Molecular; Molecular Target; Mus; Neoadjuvant Therapy; Online Systems; Pathologic; Patient-Focused Outcomes; Patient-derived xenograft models of breast cancer; Patients; Phenotype; Physiological; Prediction of Response to Therapy; Prognosis; Proteins; Recurrence; Regimen; Reproducibility; Research; Research Personnel; Resource Informatics; Resources; Science; Signal Transduction; Surrogate Endpoint; Testing; Time; Treatment Efficacy; Treatment Protocols; Work; animal data; arm; base; bioinformatics resource; candidate identification; chemotherapy; co-clinical trial; cohort; contrast enhanced; data resource; data sharing; diffusion weighted; docetaxel; exome; high resolution imaging; human data; human model; imaging biomarker; imaging modality; improved; indexing; informatics tool; innovation; interest; mRNA Expression; machine learning algorithm; magnetic resonance imaging biomarker; malignant breast neoplasm; molecular marker; novel; online resource; optimal treatments; patient derived xenograft model; personalized therapeutic; pre-clinical; preclinical trial; predicting response; predictive marker; predictive modeling; prospective; quantitative imaging; repaired; response; response biomarker; targeted treatment; tool; transcriptome sequencing; treatment response; triple-negative invasive breast carcinoma; tumor

Project Summary Triple negative breast cancer (TNBC) is a very challenging disease because it is biologically aggressive, there are no targeted therapies, and, consequently, patients have poor prognosis. Although immunotherapy is promising for treating many cancers, TNBC lacks specific molecular targets, no predictive biomarkers to chemotherapy response have yet been identified, and treatment response is difficult to evaluate using current biomarker assessments. Patient-derived xenograft (PDX) models of TNBC offer the exciting opportunity of evaluating this disease in terms of molecular features (e.g., genomic copy number, whole exome sequence, and mRNA expression) to identify candidate “omic” biomarkers that best predict the ultimate response to treatment and could provide surrogate endpoints to validate novel imaging biomarkers in co-clinical trial human trails. Moreover, emerging quantitative MRI methods, such as dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) and diffusion weighted MRI (DW-MRI), contain rich physiological signals in the images for predicting treatment response, but it is challenging to integrate both animal and human data to reliably predict the treatment response. A paradigm of “co-clinical trials” is emerging in which new treatments are evaluated in animals, and the results guide treatments in clinical trials, but there is a paucity of informatics tools and resources to enable analyses in such animal-to-human work. We believe that an informatics-based methodology that integrates molecular `omics' and imaging data will propel advances in TNBC by enabling development of machine learning models to predict the response to therapies. In order to develop research resources that will encourage consensus on how quantitative imaging methods are optimized to improve the quality of imaging results for co-clinical trials, we will leverage an ongoing co-clinical trial we are undertaking to pursue the following specific aims: (1) Identify molecular biomarkers that predict response in TNBC patient-derived xenografts (PDX); (2) Identify quantitative MRI biomarkers that predict response in TNBC patient-derived xenografts; and (3) Evaluate our informatics tools in a prospective co-clinical trial. Our proposed research is significant and innovative because it leverages advances in basic cancer biology, state-of-the-art imaging technologies, and informatics methods to develop a resource to catalyze discovery in this important disease. Our PDX-based approach will provide the cancer community with a rational, iterative, combined pre-clinical and clinical methodology and supporting data resource for making progressively more refined and personalized therapeutic regimens for TNBC patients. Our methods and tools will likely also generalize to other cancers and could, therefore, substantially benefit the care of all cancer patients.

项目摘要 三重阴性乳腺癌（TNBC）是一种非常挑战的疾病，因为它具有生物学侵略性， 尽管免疫疗法不是靶向疗法，因此，患者的预后不良。 TNBC有望治疗许多癌症，缺乏特定的分子靶标，没有预测性生物标志物 尚未确定化学疗法反应，并且很难使用电流评估治疗反应 生物标志物评估。 TNBC的患者衍生异种移植（PDX）模型提供了令人兴奋的机会 根据分子特征评估该疾病（例如，基因组拷贝数，整个外显子组序列和 mRNA表达）以识别候选的“ OMIC”生物标志物，以最能预测对治疗的最终反应 并可以提供替代端点，以验证共同临床试验中的新成像生物标志物。 此外，新兴的定量MRI方法，例如动态对比度增强的磁共振 成像（DCE-MRI）和扩散加权MRI（DW-MRI）在图像中包含丰富的生理信号 预测治疗反应，但要整合动物和人类数据以可靠地预测 治疗反应。出现了“共同临床试验”的范式，在其中评估了新的治疗方法 动物以及结果指导临床试验中的治疗方法，但信息的工具和资源很少 在这种动物到人类的工作中进行分析。我们认为，基于信息的方法论 综合分子“ OMIC”和成像数据将通过发展 机器学习模型以预测对疗法的反应。为了开发研究资源 鼓励就如何优化定量成像方法达成共识以提高成像的质量 共同临床试验的结果，我们将利用我们正在进行的持续共同临床试验来追求以下 具体目的：（1）确定可以预测TNBC患者衍生Xenographtics（PDX）反应的分子生物标志物； （2）确定预测TNBC患者衍生Xenographys中反应的定量MRI生物标志物； （3） 在一项前瞻性共同临床试验中评估我们信息丰富的工具。我们拟议的研究很重要， 创新的，因为它利用基本癌症生物学，最先进的成像技术和 提供资源来催化这种重要疾病发现的资源的信息。我们的基于PDX 方法将为癌症社区提供合理的，迭代的，临床前和临床的合并 方法和支持数据资源，以逐步精致和个性化的治疗 TNBC患者的方案。我们的方法和工具也可能会推广到其他癌症，并且可以 因此，实质上有益于所有癌症患者的护理。