ESTRO 2023 - Abstract Book

S838

Tuesday 16 May 2023

ESTRO 2023

Level I evidence-based medicine relies on randomized controlled trials designed for large population of patients. But the increasing number of clinical and biological parameters that need to be explored to achieve precision medicine makes it almost impossible to design dedicated trials. Moreover, the similarity between clinical research patients and routine care patients regarding comorbidities, severity, time before initiation of treatment and tumor characteristics has been questioned. New approaches are needed for all subpopulations of patients. Clinicians need to use all the diagnostic tools (medical imaging, blood tests and genomics) in order to decide the appropriate combination of treatments (radiotherapy, chemotherapy, targeted therapy and immunotherapy). Each patient has an individual set of molecular abnormalities responsible for their disease or correlated with treatment response and clinical outcome. The concept of tailored treatments relies on identifying and leveraging these aberrations for each patient. This shift to molecular oncology has driven cancer research in the last 25 years and has allowed significant progress in poor-prognosis diseases such as non-small cell lung cancer (through the use of EGFR inhibitors) or melanoma (through the use of immunotherapy). Unfortunately, in radiation oncology, we have not been able to leverage the same methods to better tailor our treatments. A new paradigm of data driven methodologies reusing routine healthcare data to provide decision support is emerging. Clinical decision support algorithms will be derived entirely from data and AI. Integrating such a large and heterogeneous amount of data is in itself a challenge that must be overcome before we can actually create accurate models. The objective of this presentation is to discuss how we should implement precision medicine programs in radiation oncology and describe approaches to address these challenges. Abstract Text While radiotherapy has mainly improved thanks to the improvement of imaging techniques, the development of artificial intelligence methods in imaging offers new perspectives for a radiotherapy that is more and more personalized to the patient's disease. Indeed, AI has shown great results in developing imaging tools to analyze the cellular and anatomopathological data behind the image. The applications of these new technologies seem very promising. For example, these approaches could help the clinician to identify more precisely the targets to be treated in multi-metastatic diseases, or to better define the target volumes, in particular the microscopic extensions (VCT), and the areas at higher risk of recurrence. The aim of this presentation is to discuss whether imaging and IA may help to guide clinician towards a new area of precision medicine, with histology-driven imaging biomarkers for radiation prescription. SP-1003 How to quantify the added value of automated (artificial intelligence) based RT workflows in 2030 W. van Elmpt The Netherlands Abstract Text Artificial intelligence (AI) and machine learning (ML) approaches are transforming radiation oncology by facilitating automation and optimizing workflows for efficient high quality radiation therapy administration. Examples of emerging AI technologies influencing healthcare can be found across the entire care trajectory. In particular, AI applications across the radiation treatment pathway including those found in QA and QC processes, CT and MR simulation, multimodal image fusion, automatic segmentation, synthetic CT generation, treatment planning, online and offline adaptation and image guided radiation therapy will significantly affect how Radiation Therapists work and the decisions they make. Although the efficiencies afforded by automation are appealing to Radiation therapists and promise a focus on higher level tasks and patient care, there are risks and concerns that should not be ignored. This talk will focus on the risks of automation from a Radiation therapists’ perspective and potential solutions for consideration. The importance of AI and ML curriculum for undergraduate Radiation therapy education, participation in activities prior to and in preparation for AI clinical implementation to ensure safe and clinically relevant utilization, performing end to end testing with the multidisciplinary team, providing training and education for colleagues, understanding and performing regular QA processes and upkeeping fundamental radiation therapy domain knowledge will be discussed. The role of advanced practice, and Clinical / Application specialists will also be considered. Automation is inevitable. How Radiation therapists respond is critical. SP-1002 2030: Towards histology-driven radiation prescription (imaging to pathology translation) R. Sun 1 1 Gustave Roussy Cancer Campus, Département de radiothérapie , Villejuif Cedex, France Abstract not available SP-1004 What are the risks in a fully automatic workflow? The RTT point of view C. Dickie 1 1 Princess Margaret Cancer Centre, Radiation Medicine Program, Toronto, Canada

Debate: Radiotherapy and oligometastatic prostate cancer - This house believes we are ready for a new treatment paradigm

SP-1006 For the motion A. Tree 1 1 Royal Marsden NHS Foundation Trust, Urology, Sutton, Surrey, United Kingdom

Abstract Text