Integration of Multiple Data Levels to Improve Diagnosis, Predict Treatment Response and Suggest Targets to Overcome Therapy Resistance in High-grade Serous Ovarian Cancer
试验速览
- 阶段
- 不适用
- 状态
- 招募中
- 入组人数
- 200
- 试验地点
- 2
- 主要终点
- Successful prediction of patient outcome with AI methods
研究概览
简要总结
Chemotherapy resistance is the greatest contributor to mortality in advanced cancers and severe challenges remain in finding effective treatment modalities to cancer patients with metastasized and relapsed disease. High-grade serous ovarian cancer (HGSOC) is typically diagnosed at a stage where the disease is already widely spread to the abdomen and current standard of practice treatment consists of surgery followed by platinum-taxane based chemotherapy and maintenance therapy. While 90% of HGSOC patients show no clinically detectable signs of cancer after surgery and chemotherapy, only 43% of the patients are alive five years after diagnosis because of chemoresistant cancer.
This prospective, observational trial focuses on revealing major mechanisms causing chemoresistance in HGSOG patients and derive personalized treatment regimens for chemotherapy resistant HGSOC patients. The investigators recruit newly diagnosed advanced stage HGSOC patients who are then thoroughly followed during their cancer treatment. Longitudinal sampling includes digitalized H&E stained histology slides mainly collected during routine diagnostics, fresh tumor & ascites samples for next-generation sequencing/proteomics (WGS, RNA-seq, DNA-methylation, ATAC-seq, ChIP-seq, mass cytometry, etc.) and ex vivo experiments, plasma samples for circulating tumor DNA (ctDNA) analyses. Broad range of clinical parameters such as laboratory and radiologic parameters (e.g., FDG PET/CT), given cancer treatments and their outcomes are collected. Radiomic analyses are performed to PET/CT and CT scans. Long-term patient derived organoid lines are established from fresh tumor tissues. Actionable genomic alterations are searched.
The general objective is to establish a clinically useful precision oncology approach based on multi-level data collected in longitudinal setting, and translate the most potent and validated discoveries into clinical use. DECIDER project will produce AI-powered diagnostic tools, cutting-edge software platforms for clinical decision-making, novel data analysis & integration methods, and high-throughput ex vivo drug screening approaches.
详细描述
Specific aims include:
- Develop tools and methods for personalized medicine approaches to cancer patients.
- Develop open-source visualization and interpretation software that facilitate clinical decision making via data integration and interpretation of multilevel data from cancer patients.
- Rapidly identify HGSOC patients who are likely to respond poorly to current therapies combining information on digitalized histopathology samples, genomic and clinical data with AI methods.
- Deploy validated personalized medicine treatment options using longitudinal measurement and ex vivo organoid cultures from cancer patients in clinical care.
研究设计
- 研究类型
- Interventional
- 分配方式
- Non Randomized
- 干预模型
- Parallel
- 主要目的
- Basic Science
- 盲法
- None
入排标准
- 年龄范围
- 18 Years 至 —（Adult, Older Adult）
- 性别
- Female
- 接受健康志愿者
- 否
入选标准
- •Patients with a suspected ovarian cancer diagnosis treated at the Turku University Hospital
- •Ability to understand and the willingness to sign a written informed consent document
排除标准
- •Age <18 years, too poor condition for active treatment (surgery, chemotherapy)
- •FDG PET/CT scan is not performed for patients with diabetes mellitus and poor glucose balance.
结局指标
主要结局
Successful prediction of patient outcome with AI methods
时间窗: 5 years
Proportion of patients whose disease outcome (PFS, OS) is predicted correctly with digital histopathology images, genomic data and routine laboratory values
Successful clinical translation
时间窗: 5 years
The magnitude of successful clinical translation is measured by the number of times project-derived personalized medicine has impacted patients care by application of novel and existing biomarkers and therapies.
次要结局
- Establishment of an updated version of Chemoresponse score (CRS) for measuring histological effect in tumor tissue after chemotherapy(5 years)
- Prediction of primary treatment response from tumor histology using H&E stained whole slide images and AI-based methods(5 years)
- Successful validation of potentially druggable genetic alterations(5 years)
- Successful prediction of genomic features from tumor histology(5 years)