Prof. Harald Schmidt, Chair of Pharmacology and Personalized Medicine at Maastricht University and Coordinator of the REPO4EU Project, explains how artificial intelligence could unleash a revolution in reusing existing drugs to treat diseases once thought untreatable.
Many drugs can treat diseases beyond the ones for which they were initially designed. Why aren’t healthcare systems using this potential?
The drug market is currently geared towards new, patent-protectable, and high-return-on-investment compounds. Repurposing a drug from one indication to another has so far been mostly serendipitous and of marginal relevance. The system is not designed to encourage or reward this approach, despite the enormous scientific and medical potential.
The science is solved. The remaining challenge is the lack of financial incentives for industry and healthcare providers. Repurposed drugs are often inexpensive generics, which makes them, in the current legal and reimbursement framework, unattractive from a business perspective, even if they could benefit large patient groups and save costs for our healthcare systems. Without innovative reimbursement models and regulatory frameworks that properly reflect the clinical and societal value of these therapies, the incentive to systematically explore them will remain low. Changing this mindset top down, i.e., at the policy level, is essential if we want science-driven innovation to reach patients.
A few well-known drugs already have multiple uses. Aspirin, Metformin, and diabetes drugs are most prominent. Are they exceptions, or are there many more medicines with more “powers”?
There are many more. On average, a drug on the market binds to 32 other proteins, each of which could in principle lead to a new indication. Terms like “beta-blockers” are misleading; these compounds do not only bind to beta-receptors, for example. What this tells us is that existing medicines are often far more versatile than the labels suggest, and their true potential has barely been explored systematically.
The fact that so many drugs interact with multiple biological targets means we are only scratching the surface of what they can do. If we systematically map these interactions and connect them to disease mechanisms, i.e, REPO4EU’s protein-protein interaction module concept, we could uncover an entire hidden landscape of therapeutic opportunities. This approach could even transform the way we think about pharmacology – from “one disease, one symptom-based, low-precision drug” to “one mechanistic network, two or three low-dose synergistic, high-precision, curative drugs.” The result would be a far more efficient and sustainable use of safe, existing medicines.
To make discoveries, we need real-world data, such as that from electronic health records. Do we have enough access to this kind of data?
Absolutely. In Europe, for example, we have the BBMRI biobank infrastructure and its connected data repositories. The most valuable datasets are those collected both cross-sectionally and longitudinally, without bias, and over several years. They enable us to observe time-dependent changes in patients and provide valuable insights when combined with genetic and molecular data.
Yet, access to such data is still uneven across regions and institutions, and interoperability remains a major hurdle. We need stronger frameworks for secure data sharing that respect privacy while enabling large-scale analyses across borders. Integrating real-world clinical information with molecular profiles presents not only a technical challenge but also a cultural one, requiring trust and collaboration among healthcare providers, researchers, and patients. If we succeed, the payoff will be transformative for precision, curative medicine, and our ultimate goal, prevention.
Most repurposing successes were discovered by chance. How can AI and bioinformatics change that?
Traditionally, one observed by accident that a drug unexpectedly reduced certain symptoms. However, symptom relief is not enough; it does not work for every patient, and it rarely means curing the disease.
AI and bioinformatics change the game. They help us link risk genes to diseases, recognize hidden patterns, and redefine diseases based on causal genetic risks and protein-protein interaction modules rather than just symptoms. For the first time, this approach allows us to treat diseases with precision, and ideally, even cure or prevent them.
Through projects like REPO4EU, researchers worldwide may soon gain access to our user-friendly platform, which makes cutting-edge systems medicine available to all, not just specialized bioinformaticians.
By connecting scientific discovery with all downstream steps – from intellectual property management to clinical validation - REPO4EU aims to close the current translational “valley of death” between academic output and patient benefit.
If policymakers, scientists, and patient communities can align, the future of drug repurposing could bring cures and treatments to millions of patients currently left without options. AI is speeding up research, and it may well redefine the entire landscape of what we call a disease and its therapy, keeping in mind that 60-80% of all disease risks could be kept under control with appropriate lifestyle changes.
Even when off-label uses are found, reimbursement is a barrier. What’s blocking progress here?
We do not have problems getting these new therapies patented and approved by regulators. The big hurdle is reimbursement. When we repurpose or combine drugs, they are often very cheap. But current reimbursement guidelines only accept the lowest market price for such drugs.
If these rules remain unchanged, hundreds of new, effective, and cost-saving therapies will never reach patients. To tackle this, we have initiated annual policy events in Brussels. It may sound unusual for a scientist to engage in politics, but it is essential to ensure that our research truly benefits patients.
REPO4EU is building a platform that brings together data, training, and expert support. How will it change the way researchers work?
The key benefit will be to the millions of biomedical researchers who currently have no access to state-of-the-art bioinformatics tools. We are building an easy-to-use interface that requires no coding expertise, essentially systems medicine “on speed.”
But it is more than just a website. It is a professional workspace that guides users through all subsequent steps: patenting, clinical trial design, finding collaborators, business development, and more. In short, it empowers researchers to reality-check and translate their discoveries into patient-relevant therapies.
This integrated environment will also reduce the time and cost required to move from a scientific insight to a clinical application. Instead of navigating a fragmented ecosystem of tools and stakeholders, researchers can follow a guided path from discovery to patients and product. By lowering the entry barrier for complex analyses, we hope to democratize drug repurposing and systems medicine. Ultimately, this could shift the innovation landscape from being industry-led to science- and patient-driven.
Patents and ownership issues often stall repurposing. How will your project help address these challenges?
Awareness is key. We need to engage policymakers, decision makers, patient organizations, and the public. If the current legal and regulatory frameworks do not change, hundreds of effective high-precision therapies will remain locked away from patients.
REPO4EU aims to create the critical mass and momentum necessary to place this issue firmly on the political and healthcare agendas.
Which diseases hold the biggest promise for repurposed drugs?
Essentially, all diseases that cannot currently be cured. The easiest cases are monogenetic, rare diseases, because they are in many cases directly linked to a known single molecular defect. That is why we are preparing a major global initiative to tackle untreatable rare diseases in collaboration with patient organizations, doctors, regulators, and ethics experts.
Finally, what recent advances in AI excite you most for pharmacology?
It is the speed. Tasks that used to take us months can now be completed in minutes. It feels like having a virtual team of hundreds of staff members, all working toward one shared goal. This acceleration is not just convenient; it fundamentally reshapes what is scientifically and clinically possible.
Equally exciting is the emergence of AI systems that do not just analyze data but detect patterns and thereby help to generate new hypotheses. These models can propose novel drug–disease connections, predict synergistic combinations, and even suggest optimized clinical trial designs. By automating what used to be labor-intensive and time-consuming tasks, these systems free up scientists to focus on higher-level strategic questions. This synergy between human expertise and machine intelligence will define the next era of drug therapy.
We may already have all the drugs we need!
