ESTRO 2022 - Abstract Book

S773

Abstract book

ESTRO 2022

In total, 69 patients (9.1%) experienced G ≥ 3 LT, with 18 (2.4%) experiencing more than one toxicity. Of note, G ≥ 3 LT are more common in those tumour groups (such as Chordomas and Chondrosarcomas) receiving dose-escalated treatments, and in those (such as Craniopharyngiomas and Ependymoma) commonly located in close proximity to critical organs at risk and background of multiple surgical procedures. Conclusion The results of this study indicate safety of PBT for CNS tumours, with predominantly passive scattering. Clinical outcomes from this cohort will be compared with the newest Pencil Beam Scanning technology, in the UK NHS National PBT service. Baseline toxicity assessment is essential for the correct interpretation of radiotherapy toxicities and longer follow up is needed to evaluate endpoints, such as secondary malignancies, which have a long latency period.

MO-0884 Machine learning on clinical data for mortality risk-stratification after radiotherapy for NSCLC

S. Hindocha 1 , T. Charlton 2 , K. Linton-Reid 3 , B. Hunter 1 , C. Chan 4 , M. Ahmed 1 , E. Robinson 5 , M. Orton 6 , S. Ahmad 2 , F. McDonald 1 , I. Locke 1 , D. Power 7 , M. Blackledge 8 , R. Lee 9 , E. Aboagye 3 1 The Royal Marsden NHS Foundation Trust, Lung Unit, London, United Kingdom; 2 Guy's & St Thomas' NHS Foundation Trust, Lung Unit, London, United Kingdom; 3 Imperial College London, Department of Surgery & Cancer, London, United Kingdom; 4 The Royal Marsden NHS Foundation Trust, Clinical Oncology, London, United Kingdom; 5 Institute of Cancer Research, Clinical Trials Unit, London, United Kingdom; 6 Institute of Cancer Research, Artificial Intelligence Imaging Hub, London, United Kingdom; 7 Imperial College Healthcare NHS Trust, Clinical Oncology, London, United Kingdom; 8 Institute of Cancer Research, Radiotherapy & Imaging, London, United Kingdom; 9 The Royal Marsden NHS Foundation Trust, Early Diagnosis & Detection, London, United Kingdom Purpose or Objective Surveillance is universally recommended for NSCLC patients treated with curative-intent radiotherapy. High-quality evidence to inform optimal surveillance strategies is lacking. Machine learning (ML) demonstrates promise in accurate outcome prediction for a variety of health conditions. The purpose of this study was to utilise readily available patient, tumour, and treatment data to develop, validate and externally test machine learning models for predicting overall survival (OS). Such models may allow for personalised follow-up resulting in potentially earlier detection of recurrence for high-risk patients or avoidance of unnecessary hospital visits for low-risk patients. This would have implications for patient care and healthcare resource use globally. Materials and Methods A retrospective, multicentre study of patients receiving curative-intent radiotherapy for NSCLC was undertaken. A total of 657 patients from 5 hospitals were eligible for inclusion. Data pre-processing derived 34 features (variables) for predictive