NEW YORK – A new European research project aims to deliver a cancer genome interpretation platform that can be implemented clinically across Europe. Called CGI-Clinics, the effort has an overall budget of €10 million ($10.5 million), most of it contributed by the EU. Among its objectives is to have a CE-IVD marked tool available by the end of 2027.
The Institute for Research in Biomedicine in Barcelona, Spain, known as IRB Barcelona, is leading CGI-Clinics and the further development of a research tool called Cancer Genome Interpreter (CGI) that will be piloted with various European clinical partners to prepare it for clinical use.
"We realized there was an unmet need in the field, which is the interpretation of cancer variants in the clinical setting," said Olivia Tort, international projects manager at IRB Barcelona. "We also realized that the tool was good for research, but not adapted enough for clinical use," she said. "So we decided to adapt this tool to clinical settings."
CGI has been publicly available for several years, and scientists have described it in several publications, including a 2018 paper in Genomic Medicine and a 2021 article in Nature. The platform can be used to flag tumor alterations that drive disease or that are therapeutically actionable, and relies on different computational methods, including BoostDM and IntOGen, to annotate the alterations in a tumor according to multiple levels of evidence.
"CGI was designed with more of a research focus," commented Santiago Demajo, a researcher at IRB Barcelona involved in CGI-Clinics. "We would like to go one step further and implement a better tool to interpret cancer tumor mutations, one with a clear clinical orientation," he said. According to Demajo, such a tool would feature an interface and reports tailored to clinicians rather than researchers.
CGI-Clinics commenced on Nov. 1 with the stated goal of addressing shortcomings in the current tools used to interpret cancer mutations uncovered via next-generation sequencing. Such tools tend to vary from clinic to clinic, do not manage variants of unknown significance well, and leave patients out of the loop, Demajo said. They also rely on a patchwork of databases and resources, which takes time and might lead to suboptimal clinical decisions.
"Each hospital designs its own way" of interpreting data, said Demajo, and people tend to use different databases and different tools. "One aspect of the project is to make all of the databases accessible in one tool" that everyone can use, he said.
CGI-Clinics has three phases, including setup, a validation pilot with nine clinical partners, and then a replication study across 30 European hospitals. During these studies, project members will collect data on interpretation as well as other factors, such as cost. The participants will also deploy an application called eduCGI to help patients better understand their genomic information, making it easier for them to converse with clinicians and potentially share their data for research use. The CGI-Clinics project will run through the end of October 2027.
In addition to IRB Barcelona, other participating institutions include Vall d'Hebron Institute of Oncology (VHIO) and the Center for Genomic Regulation, also based in Barcelona; Institut Gustave Roussy and the Léon Bérard Center in France; and the University of Manchester in the UK. Other partners include the Andalusian Health Service, the Girona Biomedical Research Institute, the Catalan Institute of Technology, the Althaia Foundation, the Spanish Association Against Cancer, and Alira Health in Spain. University Clinic Cologne and University Clinic Aachen in Germany are also participating, as are the European Cancer Patient Coalition, headquartered in Brussels, and the UK-based European Association for Cancer Research.
And the list is growing. Tort said that a new partner from Greece would likely join CGI-Clinics.
Within the first two years of CGI-Clinics, the consortium aims to develop a new version of CGI for clinical use, including a molecular tumor board that will provide support to oncologists, Tort said. In the later phases, researchers will compare the performance of CGI-Clinics against conventional approaches used by partners, and will also integrate data regarding time spent doing analysis, other cost savings, and patient feedback.
Ultimately, the CGI-Clinics team would like to obtain European regulatory approval for CGI, but it is unclear whether IRB Barcelona or a separate entity would lead this initiative. Alira Health, a healthcare consulting company also based in Barcelona, is taking part in CGI-Clinics, and will provide services related to evidence generation and gaining regulatory approval. A spokesperson for Alira noted that the final decision on how to make CGI available to users rests with the consortium. Tort said that the team is committed to making it as widely available to European users as possible.
Ana Vivancos, a principal investigator of VHIO's cancer genomics group, said in an email that she will lead a project within CGI-Clinics to develop a virtual molecular tumor board. Such a tool would help oncologists interpret variants using CGI, prioritize driver and targetable alterations, and consult on novel biomarkers and the availability and suitability of early phase trials.
"Hospitals in Europe are implementing NGS in their routine biomarker testing in oncology, which is very good news, but genomic tests are in a continuous evolution at a pace that may be complex for oncologists to follow," said Vivancos. Panels are also becoming larger, and more information is appearing in genomic results, she added.
To build the virtual molecular tumor board, CGI-Clinics will work with specialists throughout Europe, so that their experiences can be used to inform a tool that can assist physicians and patients. The goal, she said, is to produce a "flexible tool for oncologists to rely on and find quick, on-spot recommendations to their doubts or questions."
Should CGI-Clinics obtain a CE-IVD mark for its platform by the time it wraps up in 2027, it will continue to maintain and improve the resource, Tort noted. She said that data curation will remain a challenge, as science continuously evolves and new biomarkers are identified. "We want to make this sustainable, and it's important to be able to update it," Tort said.
Tort added that CGI-Clinics could also be useful to companies developing targeted therapies. "Eighty percent of mutations are variants of unknown significance," she said. "This is an unexplored field that is ripe for companies to test new drugs."
As the tool is adopted, it will also improve, Demajo added, as its algorithms learn from cancer patient data. He said that large datasets covering tens of thousands of patients had been used to inform the creation of CGI, and as more data pours in, its algorithm will absorb it.