Current Projetcs

Background: Treatment decisions in head and neck cancer (HNC) are increasingly complex, yet clinicians often lack timely access to institutional real-world data to support personalized care. This project addresses this gap by developing and validating an interactive dashboard on institutional historic data that enables rapid exploration of treatment options and outcomes in patients resembling the case at hand.

Objective and Hypothesis: The overarching objective is to improve clinical decision making with a transparent, user-centric tool that uses similarity-based retrieval to identify comparable historical patients and present their treatments, toxicities, and outcomes in intuitive visual formats.

Methods: We will leverage a curated multimodal dataset of approximately 1,900 HNC patients treated at Inselspital Bern. Four work packages will: (1) finalize data structures, cohort-matching variables, and outcome definitions; (2) build the interactive dashboard and backend, including similarity-based cohort retrieval and visualizations of survival, toxicities, tumor response, and radiotherapy parameters; (3) assess usability in sessions with end-users; and (4) conduct a controlled validation study in which clinicians review blinded cases with and without dashboard support. Endpoints include decision optimality, confidence, consistency, and time-to-insight.

Expected Outcome and Impact: The project will deliver a clinical tool that identifies comparable cases and supports more consistent, transparent decision making. For patients, this enables clearer counselling, stronger involvement in shared decisions, and a more equitable chance of optimal care. For clinicians and multidisciplinary meetings, the dashboard increases transparency and strengthens data-driven practice. Long term, the infrastructure provides a scalable foundation for future research and extension to other tumor types and institutions.

Dr Robert Poel is Lead for Clinical Translation of Artificial Intelligence in Radiation Oncology at the University Clinic for Radiation Oncology, Inselspital Bern. He holds a PhD in Biomedical Engineering (summa cum laude) from the University of Bern and a Master’s degree in Technical Medicine from the University of Twente, the Netherlands. His research focuses on the clinical translation of artificial intelligence and real-world data into radiation oncology, with expertise spanning clinical decision support, deep learning, radiotherapy quality assurance, and data infrastructure. Dr Poel has led the development and clinical evaluation of AI-driven solutions for radiotherapy and has authored numerous peer-reviewed publications in medical imaging and radiation oncology. His current work aims to integrate transparent, data-driven tools into routine clinical practice to improve personalized cancer care.

Bladder cancer (BC) is a significant public health problem worldwide, in terms of prevalence, mortality, clinical management and cost. A big proportion of patients progress from non–muscle-invasive BC to muscle invasive BC and respond poorly to current therapies, including immune checkpoint inhibitors and tyrosine kinase inhibitors. Understanding the molecular mechanisms driving progression and therapeutic resistance, and identifying robust biomarkers for patient stratification and monitoring, remains a critical unmet need.

Our preliminary analyses revealed a 16-gene program, where PDGFRβ plays a central role, that is associated with tumor aggressiveness. We identified a tumor-intrinsic role for PDGFRβ: its elevated expression in cancer cells correlates with advanced stage, disease progression, and poor prognosis. Importantly, genetic or pharmacological inhibition of PDGFRβ reduces tumor invasiveness, controls tumor growth, limits metastatic spread, and shifts the tumor microenvironment toward an immune-inflamed state. These findings position PDGFRβ as a potential regulator of BC progression and therapy resistance.

This project comprehensively dissects PDGFRβ signalling across the full clinical spectrum of BC through an integrated multi-omics approach. Single-cell spatial transcriptomics and proteomics will map PDGFRβ expressing tumor ecosystems in patients’ samples. Large independent cohorts will validate its predictive and prognostic value, including in anti-PD-1–treated patients. Complementary liquid biopsies will support the exploratory development of a non-invasive PDGFRβ-linked biomarker panel for disease monitoring and therapy orientation.

By combining cutting-edge spatial technologies, liquid biomarker profiling, well-annotated multi-centric patient cohorts, and functional validation, this work aims to define PDGFRβ as a precision oncology target and deliver actionable tools to counteract BC progression and improve immunotherapy responsiveness.

Originally from Rabat, Morocco, Dr Ahanidi is a promising young researcher specialising in cancer biology. After earning a Master’s degree and a PhD in Molecular biology from the Faculty of Medicine in Rabat, she pursued advanced research training at the Geisel School of Medicine at Dartmouth in New Hampshire, USA. In 2022, she joined the Department of Pathology and Immunology at the Faculty of Medicine, University of Geneva. Her research focuses on unravelling the molecular mechanisms driving bladder cancer progression and identifying innovative therapeutic strategies to improve patient outcomes.