The webinar, “Advancing Immuno-Oncology – Bridging Into Biomarkers, Therapy, and Translation,” brought together leading voices from academia, biotech and pharma to explore how biomarkers, AI and next-generation therapeutic strategies are reshaping cancer drug discovery and development.
The session featured an interview with Professor Sophia N. Karagiannis, Professor of Translational Cancer Immunology and Immunotherapy at St John’s Institute of Dermatology, School of Basic & Medical Biosciences, King’s College London. The panel discussion was moderated by David Krige, Chief Development Officer at Accession Therapeutics, and included panellists Lilly Wollman, CEO of Synteny AI; Professor Sophia N. Karagiannis; and Vaios Karanikas, Translational Biomarker Sciences Group Head, Expert Biomarker Experimental Medicine Leader, Early Development Oncology, Roche Pharma Research and Early Development, Roche Innovation Center Zurich.
Professor Karagiannis opened the discussion by outlining how her lab is working to connect cancer biology, immune mechanisms and clinical outcomes. Her research focuses on understanding how cancer modulates immune sentinels, particularly B cells, myeloid cells and natural killer cells, and how these insights can inform the design of next-generation antibody therapeutics. As she explained, effective therapeutic development must consider the wider immune landscape because “cancer is systemic disease.”
A key theme of the webinar was the need to move beyond a narrow, T cell-centric view of cancer immunity. Professor Karagiannis highlighted the importance of B cells, antibodies, macrophages and other immune populations in shaping treatment response. Her lab integrates long-read antibody sequencing, proteomics and functional immunoassays to identify antibody signatures that may predict disease progression, treatment response or toxicity. These approaches are particularly important in areas such as melanoma, where better biomarkers are needed to understand which patients are most likely to benefit from immunotherapy.
From a pharma perspective, Vaios Karanikas of Roche emphasised the importance of understanding cancer as a systemic disease requiring complementary therapeutic strategies. “You can’t battle it with one single modality,” he said. He outlined several key therapeutic pillars being explored in early clinical development, including T cell engagers, antibody drug conjugates, cancer cell dependency approaches and next-generation immunotherapies designed to engage multiple arms of the immune system.
Karanikas also stressed the central role of biomarkers in translating biology into clinical action. “We want to understand for every drug that we bring into the clinic the biology of the drug,” he said. For Roche, this means using biomarker strategies to understand not only why a drug works, but also why it may fail in some settings, helping identify “the right patient for the right drug at the right time.”
The panel also explored the growing role of AI in immuno-oncology. Lilly Wollman, CEO of Synteny AI, described how generative models are helping the field move from screening-based discovery toward design-based drug development. By training models on high-throughput biological data, AI may support the design of more precise TCRs, antibodies and other biologics with improved specificity, affinity and reduced off-target effects. “It really starts with the data,” she said, noting that high-quality, diverse datasets are essential for training useful models.
Wollman also discussed the potential of AI to support patient stratification by learning immune signatures from pre- and post-treatment samples. In her view, understanding the immune system before treatment is critical, asking: “What is the immune system of that patient prior to treatment?” Such insights could help identify which patients are most likely to respond to checkpoint inhibitors, T cell engagers or other immune-based therapies.
As moderator, David Krige brought a clinical development and biotech perspective to the conversation. He highlighted Accession Therapeutics’ work in oncolytic viruses, describing them as attractive because “they are a combination therapy in one.” These therapies have the potential to directly lyse tumour cells, release tumour antigens and stimulate immune responses, while also enabling the delivery of potent transgenes specifically within tumours.
Krige also reflected on the practical realities of early-stage biotech development, particularly when working with limited resources and patient samples. “We think about this deeply, and then the next thing we think about is, right, so what are we going to do, and what are we not going to do, because we can’t do everything,” he said. For smaller companies, prioritising the most informative biomarker and proof-of-concept studies is essential.
Despite the promise of new modalities and AI-enabled approaches, the speakers acknowledged persistent challenges. Translating biology into clinical action requires robust models, high-quality patient samples, longitudinal immune profiling and strong collaboration between academia, biotech, clinicians and pharma. Professor Karagiannis noted that selecting the right models remains a major issue, while Wollman highlighted the challenge of building teams that can combine AI and biology effectively.
Looking ahead, the panel agreed that the future of immuno-oncology will depend on closer integration of biomarkers, immune profiling, tumour biology, therapeutic engineering and AI. By linking biomarkers directly to treatment selection and therapeutic design, the field can move closer to truly personalised cancer care.
Ultimately, the webinar underscored that progress in immuno-oncology will not be driven by a single technology or modality. Instead, it will come from connecting biology, biomarkers and therapeutic design in ways that make treatment more precise, personalised and clinically meaningful.
