High Grade Ovarian Serous Adenocarcinoma, High Grade Serous Carcinoma
Conditions
Keywords
chemoresistance, personalized medicine, WGS, ctDNA, digital pathology, FDG PET/CT, bioinformatics, AI, ovarian cancer, tumor evolution, artificial intelligence, organoid, DNA methylation, radiomics
Brief summary
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.
Detailed description
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.
Interventions
Sponsors
University of Helsinki
Turku University Hospital
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
BASIC_SCIENCE
Masking
NONE
Eligibility
Sex/Gender
FEMALE
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* 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
Exclusion criteria
* 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.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Successful clinical translation | 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. |
| Successful prediction of patient outcome with AI methods | 5 years | Proportion of patients whose disease outcome (PFS, OS) is predicted correctly with digital histopathology images, genomic data and routine laboratory values |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Successful validation of potentially druggable genetic alterations | 5 years | Number of potentially druggable genetic alterations found and validated with in-vitro methods |
| Successful prediction of genomic features from tumor histology | 5 years | Number of genomic features that can be successfully recognized from tumor histology |
| Prediction of primary treatment response from tumor histology using H&E stained whole slide images and AI-based methods | 5 years | Number of patients whose outcome (primary therapy outcome, PFS) is predicted correctly |
| Establishment of an updated version of Chemoresponse score (CRS) for measuring histological effect in tumor tissue after chemotherapy | 5 years | Predictive power of the updated CRS at interval surgery is compared with traditional CRS |
Countries
Finland
Contacts
Primary ContactJohanna Hynninen
Backup ContactSampsa Hautaniemi
Outcome results
None listed