Osteoarthritis (OA) occurs when the cartilage and other tissues in our joints deteriorate, resulting in pain, stiffness, and disability. It already affects more than 500 million people globally, and this number is expected to rise sharply in the next decades as populations age.
There is currently no cure for osteoarthritis, and efforts to develop treatments are hampered by several factors. These include the heterogeneous nature of the disease (there may be several different subtypes of osteoarthritis), the slow progression of the disease, our limited understanding of the underlying causes of the disease, and a lack of adequate tools to measure meaningful changes in the disease in the context of a clinical trial.
Counting on cutting-edge artificial intelligence and machine learning
The aim of PROBE is to use a big data approach, powered by cutting-edge artificial intelligence (AI) and machine learning (ML), to advance the care of people living with knee osteoarthritis.
At the heart of the project will be a federated database and a federated learning platform. In a federated network, sensitive patient data is stored locally behind secure firewalls and does not need to be transferred elsewhere. Machine learning models are trained on this locally stored data, allowing knowledge to be generated while safeguarding patient confidentiality. The resulting anonymised readouts can be transferred and shared as needed.
A focus on knee osteoarthritis
The PROBE team will leverage multiple datasets covering more than 70 million people from across the EU and beyond. Knee osteoarthritis is the main focus, as it is responsible for the bulk of the total and individual healthcare burden of osteoarthritis, and most available data on osteoarthritis is on the knee. However, based on the insights gathered for knee osteoarthritis, the team will develop a strategic roadmap for expanding the PROBE approach to other joints.
The project will harmonise the data and use artificial intelligence / machine learning approaches to enable the identification of subgroups of patients who could benefit from specific treatment approaches. Another project goal is the identification of relevant clinical trial strategies and ‘endpoints’, i.e. measures of changes in the disease that could be used to assess the results of clinical trials in knee osteoarthritis.
“Big Data analytics in this public-private PROBE consortium will hopefully change how treatments can be developed for OA, a highly prevalent disease with enormous societal impact, by pivoting towards personalised approaches with AI-based phenotyping and prediction modelling leveraging on innovative digital twin approaches,” said PROBE co-coordinator and industry lead Matthias Schieker of Novartis.
The insights gathered in the project will feed into a tool to support shared decision-making for patients and healthcare providers when discussing treatment strategies and planning care.
Ultimately, PROBE aims to accelerate the development of effective treatments for knee osteoarthritis treatments, thereby improving the lives of millions of patients.
“A joint effort by the key players in osteoarthritis research”
“We are proud to coordinate this consortium, working together to achieve a significant advancement in osteoarthritis research,” said PROBE coordinator and academic lead Sita Bierma-Zeinstra of Erasmus Universitair Medisch Centrum Rotterdam. “This project is not merely a standalone initiative with a few partners, but a joint effort by the key players in osteoarthritis research. It represents a forward-looking vision for the future of osteoarthritis research, aiming to improve treatment and long-term OA outcomes.”