NEW YORK – A consortium led by the University of Tartu and Tartu University Hospital in Estonia has been awarded €30 million ($33 million) to create a personalized medicine research and development center of excellence.
The six-year project, called TeamPerMed, involves partners in Finland and the Netherlands, and is being supported by the EU and the Estonian government, each of which will contribute €15 million.
TeamPerMed builds on decades of work undertaken at the university centered on the Estonian Biobank. The biorepository already has genotyped 210,000 participants, mostly using the Illumina Global Screening Array, and partnered in 2015 with the Broad Institute to sequence the whole genomes of 2,500 Estonians.
According to Mait Metspalu, the principal investigator of TeamPerMed and director of the University of Tartu's Institute for Genomics, the consortium intends to perform whole-genome sequencing of 10,000 individuals to build out its reference genome set.
Metspalu noted that the generation of such data will spur additional research and, potentially, new risk scores that can eventually be implemented in a clinical setting through the TeamPerMed Center of Excellence.
"This is not a project with an end," said Metspalu. "This is essentially seed money to create a new center that will become sustainable, that attracts funding both from private and public sources, and will develop research and innovation," he said. Metspalu added that the center will also play an advisory role in the future, providing valuable information to policymakers about how to best provide oversight of new clinical innovations.
The EU is supporting TeamPerMed through a "teaming for excellence" funding action through Horizon Europe. A component of such actions is partnering with other EU institutions in order to improve Europe's overall competitiveness. As such, TeamPerMed includes as partners the Institute for Molecular Medicine Finland (FIMM) at the University of Helsinki, and Erasmus University Medical Center in Rotterdam. Both parties will receive €3 million apiece for their role in TeamPerMed, Metspalu noted, and each brings different resources and expertise to the table.
FIMM has been coordinating FinnGen, a six-year-old effort to collect genotyping and health registry data on 500,000 people in Finland. According to its website, it is well on its way to achieving that goal by this year, with data currently available for 429,200 people. Numerous pharmaceutical companies have also been supporting the FinnGen project, including AbbVie, AstraZeneca, Biogen, Celgene, Genentech, Merck, and Pfizer.
According to Metspalu, TeamPerMed's Finnish collaborators therefore have experience working both with industry and researchers in the same setting. He noted that the University of Helsinki also has a talented computer science team that will work with Estonian leads on the TeamPerMed effort.
The specialty of Erasmus University Medical Center is in running clinical trials and carrying out clinical research, Metspalu said, both of which will be part of the new center of excellence's activities. "We can use their expertise here," he noted, "but we are sharing expertise," he said. "This is an exchange."
The new TeamPerMed CoE will mostly be a virtual center, combining existing resources from the university, which manages the Estonian Biobank, and the nearby Tartu University Hospital. However, it will also acquire more space to operate, and will be hiring new team members soon. Much of the €30 million budget, therefore, will be spent on salaries over the next six years.
The vision of the center is to integrate expertise in genetics, clinical medicine, public health, and socioeconomic analysis. With such a framework, it will be able to translate genomic and electronic health data into personalized medicine tools, fulfilling the vision of the Estonian Genome Project, which commenced in 2000 and led to the creation of the Estonian Biobank. As data and new tools become available, the goal of the investigators behind TeamPerMed is to test interventions in individuals in the Estonian healthcare system to prevent or delay disease onset.
In particular, the center will pilot the use of new polygenic risk scores, Metspalu said. "It's not always clear what to do with a risk score," he said. "Should you give the score to people, or fold it into national screening programs so that you identify people at risk via the screening program?" As an example, he said women could be assessed using such scores for their risk of developing breast cancer, and those with higher scores could undergo routine screening at an earlier age.
This is where the need for the 10,000 additional reference genomes is necessary, as previous sequencing efforts focused on a more homogenous subsection of the Estonian population. "You can't say that you have a service in Estonia that only works for people with a specific genetic ancestry, and we can't offer it to you," said Metspalu. "It would be unethical, to put it mildly."
He said that some of the €30 million budget will go toward supporting whole-genome sequencing of this more diverse cohort of individuals. The investigators have not selected a platform to run the whole-genome sequencing, but Metspalu said that unlike in the past, the sequencing would be carried out within Estonia.
The main data source for TeamPerMed, however, will continue to be genotyping array data, Metspalu noted, for multiple reasons, one being its lower cost — about $50 per analysis — but also because of the storage costs associated with maintaining and managing sequencing data.
"Sequencing whole genomes is getting to the price where it can be used in everyday clinical practice," said Lili Milani, professor of pharmacogenomics at the Estonian Genome Center, "but it's still a lot of data." According to Milani, by sequencing 10,000 whole genomes, the investigators involved in the TeamPerMed Center of Excellence will be able to improve the imputation accuracy for microarray data.
Milani pointed out that the previously sequenced 2,500 reference genomes had yielded numerous new discoveries, for pharmacogenetics, as an example. But even more data is needed to better understand variability in drug response, she said. Milani has for years been working with biobank data to develop risk scores relevant to drug metabolism and make those findings available to patients via the country's electronic health records system.
She characterized these previous efforts around risk scores as preliminary work, setting the stage for the activities of the new Center of Excellence. "So far we have been doing this in bits and pieces," said Milani. "We had a few grants to run clinical studies implementing polygenic risk scores, but it kind of ended there," she said. "Our arms were too short to actually reach people."
With the new funding and genomes, discovery will continue, she pledged, supported by new machine-learning tools, AI models, and phenotypic data. "It's a cycle of more information coming in, making better models, and finding ways of implementing these clinically," she said.
And it won't only be the university researchers who will be developing new scores. According to Metspalu, one goal of the center is to support an ecosystem of companies that could commercialize scores developed with the help of the Center of Excellence. One such firm, Antegenes, already exists and has launched a series of cancer risk prediction tests in Europe.
But there will be more. Milani noted that the university recently set up UniTartu Ventures, a university-owned holding company that can provide access to university IP or data licenses in return for equity in startups.
Pilot studies and data standards
Over the next six years, TeamPerMed will carry out several studies to trial the operations of the new center.
In the first pilot study, the investigators will validate the ability of AI-based tools to predict the risk of cardiovascular disease. Here, they will draw on existing health data used to assess patients for such risk, including family history data, smoking status, diabetes, and body mass index, as well as polygenic risk scores for predicting disease. They will then apply the scores in about 50,000 biobank participants with data collected in 2012, to see if the scores could accurately predict who developed disease and therefore could be used to identify high-risk, pre-symptomatic individuals most likely to benefit from early screening and intervention.
In a second study, the investigators will validate their ability to run hospital-based clinical trials on public interventions based on genomic and personalized risk assessment tools. To do this, they will assess the effectiveness of aggressive statin therapy in mitigating coronary artery disease when prescribed to young individuals determined to be of higher and medium risk. In the study, they will compare the outcomes of patients who receive lifestyle counseling based on their risk assessment results with those who receive standard treatment. The investigators will assess if the patients' disease is reduced in those who receive lifestyle counseling, and will also study the cost-effectiveness of such interventions.
The third pilot aims to validate the ability of the center to run primary-care clinical trials to assess the ability of pharmacogenetic data to guide treatment. They will do this by assessing whether making pharmacogenetic data results in increased drug adherence by patients prescribed statins who are at medium to high risk of cardiovascular disease. Again, the study will look at a cohort of patients who do receive information on risk versus those who do not under the current standard of care. The investigators believe that pharmacogenetic testing will increase patient adherence to statin therapy.
One additional project supported by the new grant involves standardizing Estonian healthcare data. Metspalu underscored that to make the TeamPerMed Center of Excellence's activities a success, phenotyping information must be recorded in a uniform manner, something that is still not done.
Researchers will therefore be working on standardizing Estonian health data according to the Observational Health Data Sciences and Informatics' Observational Medical Outcomes Partnership (OMOP) common data model and developing data-analysis solutions, according to Jaak Vilo, a professor of bioinformatics at the University of Tartu.
Vilo noted in an email that Erasmus University Medical Center is "one of the leaders in developing this ecosystem" across Europe, such as through the European Health Data Evidence Network, an Innovative Medicines Initiative-backed project that commenced in 2018 that aims to build a large-scale, federated network of European data sources aligned to a common data model.
Vilo also said that the University of Helsinki will work with Estonian researchers to "look into the algorithmic and privacy aspects of managing and analyzing large DNA and health datasets."
All of this work will attempt to harness the potential of the Estonian Biobank, said Vilo, to "achieve deep phenotyping and linking [that data] to the underlying genetics of individuals."