Professor Ryser, how did you get into cancer research?

My career path is quite unusual. I started with a degree in physics at EPFL, then did a PhD in mathematics at McGill – with a two-part thesis, one purely theoretical and the other applied to bone cell biology. I was particularly fascinated by the interplay between mathematical modelling and laboratory experimentation. A postdoc at Duke University then led me into mathematical applications in medicine, specifically cancer. I continued this work at Duke, first as an assistant professor, then associate professor, with a joint appointment between Population Health Sciences (in the School of Medicine) and Mathematics.

This background feeds into your multi-scale and multimodal approach to research...

Exactly. The biological, clinical and population levels are deeply intertwined. I try to bridge these worlds that too often stay separate. Getting ideas to flow between these scales opens up much more integrated approaches. And modelling comes into play at all levels. Cancer starts with a faulty cell that multiplies, accumulates mutations and interacts with its surroundings – dynamics we can model. When linked to molecular data from actual human tumours, these models provide a clearer picture of how the disease really progresses, which points towards better prevention and treatment. At the population level, mathematical models of tumour progression help weigh up the benefits and risks of different screening and early detection strategies, providing policymakers with essential support to develop public health policies.  
My initial discussions with cancer specialists at the HUG and UNIGE have been extremely stimulating, particularly with regard to the idea of combining clinical, imaging and molecular data to offer patients tailored care.

The RGT is at the interface between research and public health. How do you balance these two missions?

I do indeed wear two hats here at UNIGE. On the one hand, I have an academic role: building a research group focused on early cancer detection and teaching. On the other, I have an operational role at the Geneva Cancer Registry, which has a public service mission—by law, the canton must collect data on every cancer case. We have a fantastic team that processes thousands of documents to structure this data, which then goes to the National Cancer Registry (ONEC) for monitoring and research. For me, these two functions are closely linked and let us develop a truly multi-level approach to cancer. Having such diverse expertise under one roof at the RGT is a rare opportunity to push knowledge forward whilst staying rooted in Geneva's context and the needs of the local population.

What are your plans for the RGT?

There are plenty of projects ahead, both in research and teaching, as well as the RGT's public service mission. To this end, we are starting by expanding the team – we are currently recruiting a co-director for the RGT, at assistant or associate professor level. The aim is to build a larger, more visible research group around the Registry, strengthening our academic ties, particularly with the Institute of Global Health, which the RGT has historically been attached to.

Second, we want to strengthen collaborations with other public health stakeholders and clinical services in the canton. By cross-referencing our epidemiological database with other data sources, we have a real opportunity to improve health outcomes for Geneva's population. Collaborations are already in the works with the Geneva Cancer Screening Foundation and research groups in Precision Oncology and Pathology at the Faculty of Medicine and HUG.

At the same time, we will be working on our data abstraction processes to improve their efficiency and, crucially, to unlock their potential for research. While the RGT is modest in size, it stands out for the exceptional quality of its data. Our coders are detectives, hunting down relevant information in complex and heterogeneous reports. The next step is to develop AI tools to support the coding team in their work. With tens of thousands of cases recorded over nearly 55 years, the RGT data also gives us an excellent basis for training new algorithms.

In terms of teaching, what would you like to implement?

I have already taught a course called 'Math and Medicine' at Duke University, focusing on applying quantitative tools to complex biomedical problems – I'd love to offer something similar here. This course is designed for medical students without heavy quantitative training. The goal is not to master all the mathematical details, but to understand and critically evaluate the approaches being applied. I use a very interactive, hands-on teaching style, giving students a safe space to step outside their comfort zone and learn to navigate this complex science.

The second topic I am really keen on is oral communication in science: how do you adapt to your audience and get concepts across in simple terms, without jargon? These are incredibly useful skills that are still rarely taught.

Your inaugural lecture on 16 December will focus on overdiagnosis. What is it, exactly?

Like any medical test, cancer screening involves balancing benefits and risks. The benefit: catching a tumour early means less invasive treatment and a reduction in mortality. But there are downsides too, including false positives with all the stress and unnecessary workup they trigger. The most significant risk of screening, however, is overdiagnosis, defined as the detection of a malignant tumour that, left alone, would never have caused problems. For instance, the detection of a very slow-growing tumour in an elderly patient who'll die of another cause before cancer symptoms even appear. Once an overdiagnosed tumour is detected, it is treated like any other cancer—exposing the patient to side effects, but without any real benefit. The phenomenon of overdiagnoses poses several challenges. Since we cannot observe overdiagnosis directly, how can the risk of overdiagnosis be estimated? How can we identify patients at risk of overdiagnosis, and how can we scale back their treatment appropriately?

In your opinion, what are the major public health challenges in cancer?

Prevention and screening remain essential, especially with multi-cancer early detection tests just around the corner. With these tests, several cancers can be detected early based on a simple blood test. But then what? What should be done after a positive test? What clinical pathway should be offered to confirm the diagnosis while avoiding an unnecessary cascade of tests? What are the implications for individual and public health, and for healthcare costs? And how do we ensure everyone has equal access?

Another issue is the rising incidence of certain cancers in young adults. Whilst research has focused mainly on risk factors, we do not really understand how much of this increase is linked to changes in early detection practices. Better distinguishing between the effect of screening and a genuine rise in risk will be crucial for adapting our healthcare system to new epidemiological realities.