The DECIDER project started in February 2021 and will end in July 2026. The main results of the past 54 months of the project are detailed in the following.
The basis for all research in the DECIDER project is the prospective collection of tissue samples of HGSOC patients. At the Turku University Hospital in Finland, we have collected a very large, high-quality, longitudinal dataset of tissue samples, body fluids, and relevant clinical data from >400 HGSOC patients. This cohort includes patients who were recruited before February 2021 in the HERCULES study and received follow-up in the DECIDER project. The collected tissues are prepared for 1) tissue dissociation and processing as organoids and tumoroids for ex vivo validation, 2) analysis of digitalised haematoxylin and eosin (H&E) stained histopathological samples, and 3) DNA or RNA extraction and subsequent sequencing.
Whole genome sequencing data, as well as information about gene expression (RNA sequencing), are available for almost all included patients. Furthermore, we examine DNA methylation and circulating tumour DNA in plasma samples. We analyse the sequenced data to reveal changes in the patients’ genomes. This provides insight into how the tumour genome changes in response to therapy. After analysing the sequencing results, critical information is reported back to the patients’ treating clinicians. We hold regular virtual molecular tumour board (vMTB) meetings, focusing particularly on interpreting druggable targets identified in genomic data and validating them at the RNA expression level.
During the third reporting period, we further developed and validated our deep learning–based AI models to predict the chemotherapy response of HGSOC patients based on histopathological images of tumour tissues, as well as sequencing results. These tools are essential for the early identification of patients who are unlikely to benefit from standard therapy.
We have successfully grown 54 long-term organoid lines from patient samples collected in the DECIDER project. Those organoids are available for validation studies to the consortium and are the major ex vivo tool to evaluate the repurposing of drugs that were identified to be potentially usable as targets to treat resistant HGSOC.
The DECIDER patient cohort is not only large, but also longitudinal and multimodal. The open-source AI-based software Oncodash was developed to visualise all these clinical, sequencing, and histopathological data, as well as results from predictive models, and facilitate clinical decision-making based on the DECIDER findings. We furthermore published the Oncodash App Store, which serves as a centralised marketplace for sharing and distributing oncology-related software tools for the scientific community. The successful integration of DECIDER's tools into clinical workflows marks a significant achievement and is a crucial step toward personalised medicine in the treatment of HGSOC.
