Hercules Project


The HERCULES project has been divided into four subprojects:

1. Comprehensive characterisation and modelling of the spatial and temporal landscape of tumour cell subpopulations during disease progression and therapy using prospective collection of multiple subsections of freshly acquired tumours

2. Identification of genetic and transcriptomic biomarkers for tumour cell subpopulations, in particular drug resistant cell subpopulations, using mass cytometry and single-cell deep sequencing

3. Development and application of integrative computational tools that will predict efficient combination therapy modalities among molecularly-targeted anticancer compounds that are able to kill the identified subpopulations in cell cultures and in vivo

4. Development and validation of a marketable prototype biomarker kit for predicting high-grade serous ovarian cancer patients’ response to combinatorial therapeutic modalities based on a formalin-fixed, paraffin-embedded (FFPE) sample from a tumour

The project started on January 1, 2016, and its main results will be posted here and published in international peer reviewed journals.

HERCULES Data Management Plan

The latest update of the HERCULES Data Management Plan can be downloaded from here

Summary of the third periodic report 1.1.2019-30.6.2020

Summary of the context and overall objectives of the project

More than 150,000 women die of ovarian cancer every year. In the HERCULES project we aim at overcoming drug resistance in high-grade serous ovarian cancer, which is the most common and lethal subtype of ovarian cancer. Typically, the patients respond well to the platinum-based first- line chemotherapy, but often the disease becomes increasingly resistant to the treatments, leading to progressive disease and death. The mechanism of chemotherapy resistance is poorly understood, and we lack effective treatment options for patients with chemotherapy resistant disease. The major objective of the HERCULES project is to develop effective therapy options for ovarian cancer patients.

Ovarian cancers contain millions of cancer cells that have acquired a numerous and diverse set changes in their genome, such as mutations and copy number variations. Some genetic aberrations may give a cancer cell the ability to resist chemotherapy. This means that even though the cancer drug kills most cancer cells, a small fraction of resistant cells has a growth advantage and the following chemotherapy cycles are less effective. In HERCULES, we use state-of-the-art measurement technologies to obtain information from cancer samples collected at surgery before and after the chemotherapy to reveal genetic aberrations that cause chemotherapy resistance. Major part of the project is dedicated to the analysis of data from tumour samples to find biomarkers that characterize chemotherapy resistant cells. This is followed by testing combinations of drugs in patient-derived cancer cell cultures. We will use this information to identify drugs or drug combinations that would be most efficient against the chemoresistant cells.

Work performed from the beginning of the project and main results achieved so far

The HERCULES project started on January 2016 and the main results so far are 1) establishment of infrastructure where patient samples from surgery are processed and delivered to project partners for experiments, 2) functional data analysis pipelines for all data levels obtained in the project, 3) novel methods and mathematical models that enable the analysis and interpretation of large-scale data from HERCULES, 4) a database for clinical data for efficient clinical data analysis and integration and 5) biomarkers that can be used to guide treatment for a subset of the HGSOC patients.

Currently, we have recruited more than 170 high-grade serous ovarian cancer patients in the Hospital District of Southwest Finland. The number of tissue samples collected from these patients during surgery has varied from 2 to 21 per patient, and in total over a 1000 samples from 12 different anatomical sites have been collected. All the samples fulfilling quality standards have been processed for further analyses performed by the project partners. In order to connect the information from the analysis of the samples to clinical information to know how well the patient responded to the treatments, we have compiled a database where we store the clinical information relevant for the research project. The clinicians working for the project can then export clinical information concerning the samples from the database in a pseudonymized format that allows other researchers in the project to use it in their analyses.

The samples are first processed at University of Turku, where the researchers have optimised the methods to deliver as good quality samples for further experiments and analysis as possible. In order to obtain in-depth information on the tumour cells, we are using the latest sequencing and mass cytometry technologies. To identify and study the different cells that make up the tumours, we have been analysing the DNA (genome), RNA (expression of the genes), circulating tumour DNA (ctDNA; tumour DNA fragment in blood) at University of Helsinki (Finland), and setting up new experimental methods to study the regulation of genes on cell cultures at Cambridge University (UK). Measuring protein levels of important markers both on the surface and inside the cells, which has been done at the Istituto Superiore di Sanita (Italy), is essential for identifying different cell populations. A panel of markers that can be used to identify different cell populations from tumour samples has been finalised. At the Institute of Molecular Medicine (FIMM), University of Helsinki, researchers have been testing the sensitivity or resistance of cell populations from patient tumour cell cultures to different drugs. New computational methods and models for the analysis and interpretation of the large amounts of data produced in the project have been developed at University of Helsinki and Institute Pasteur (France). Preliminary biomarkers for predicting patient response have been tested in biobank samples from Finland at University of Turku and from Italy at University of Trieste to analyse their ability to predict disease progression. As we are also developing a test for predicting patient response for chemotherapy, it is important to make sure that the test can be used in diagnostic samples, which are stored correspondingly as the biobank samples. The SME partner AB Analitica (Italy), in charge of developing the test, has optimized methods for DNA extraction from such diagnostic samples, and development of the prototype test is progressing well.

Progress beyond the state of the art, expected results until the end of the project and potential impact

Measurement technologies, especially single-cell technologies, used in this project are state-of-the-art, and, when combined with our sample logistic and data analysis expertise, allow for unprecedented opportunities to identify chemotherapy resistant cells and therapies to kill them. Currently, chemoresistant cell characteristics are to a large degree unknown, and the ways to reliably identify them from a patient sample would help diagnostics, prognostication, as well as therapy decisions. Moreover, currently there are very few if any effective therapy options for patients with chemotherapy resistant disease and more efficient and targeted treatment of patients would reduce the number of expensive but inefficient and toxic treatments, hence improving both patient survival, wellbeing and savings in health care costs. When the tumours can be classified into more specific subtypes, it will also make drug development more efficient and cost effective, as the drugs can be targeted to the right patient groups in clinical trials, thus benefiting also the pharmaceutical industry and academic driven drug development projects. Results stemming from this project have been already used to facilitate clinical decision making, and we expect that by the end of the project there will be more such results.