ESTRO 2022 - Abstract Book

S720

Abstract book

ESTRO 2022

Regarding fibrosis, a statistically significant association is found, at M0, with age at diagnosis (p <0.018), obesity (OR: 1.44 p <0.001), tobacco (OR: 1.4 p <0.008), and the use of boost OR: 1.61 [1.24; 2.11] (p <0.001). Only obesity and the type of surgery received by the patient remained statistically significant at M12 and M36. The evolution of the principal skin toxicities is given in figure 1. Obesity and age at diagnosis represented at M12 and M36 a risk associated with the onset of telangiectasias. Conclusion In this study we identified several risk factors for acute and late skin toxicity such as obesity in the occurrence of skin erythema, fibrosis or telangiectasia. The use of a boost was mainly related to the occurrence of fibrosis while the use of IMRT-type technique decreased the occurrence of skin erythema (radiodermatitis). The knowledge of its predictive factors allows a personalized management of the patient by adapting our treatments and our monitoring according to these different factors. A. Cicchetti 1 , E. La Rocca 2 , M.C. De Santis 3 , P. Seibold 4 , D. Azria 5 , D. De Ruysscher 6 , R. Valdagni 7 , A.M. Dunning 8 , R. Elliot 9 , S. Gutiérrez-Enríquez 10 , M. Lambrecht 11 , E. Sperk 12 , T. Rancati 1 , T. Rattay 13 , B. Rosenstein 14 , C. Talbot 15 , A. Vega 16 , L. Veldeman 17 , A. Webb 18 , J. Chang-Claude 19 , C. West 20 1 Fondazione IRCCS Istituto Nazionale dei Tumori di Milano, Prostate Cancer Program, Milan, Italy; 2 Fondazione IRCCS Istituto dei Tumori, Radiation Oncology, Milan, Italy; 3 Fondazione IRCCS Istituto Nazionale dei Tumori, Radiation Oncology, MIlan, Italy; 4 German Cancer Research Center (DKFZ), Division of Cancer Epidemiology, Heidelberg, Germany; 5 Montpellier Cancer Institute, Radiation Oncology, Montpellier, France; 6 Maastricht University Medical Center, Radiation Oncology (Maastro), Maastricht, The Netherlands; 7 Università degli Studi di Milano, Oncology and Hemato-oncology, Milan, Italy; 8 University of Cambridge, Strangeways Research Labs, Cambridge, United Kingdom; 9 University of Manchester, Manchester Accademie Health Science Centre, Manchester, United Kingdom; 10 Vall d'Hebron Institute of Oncology (VHIO), Hereditary Cancer Genetics Group, Barcelona, Spain; 11 University Hospitals Leuven, Radiation Oncology, Leuven, Belgium; 12 Universitätsmedizin Mannheim, Medical Faculty, Mannheim, Germany; 13 University of Leicester, Cancer Research Centre, Leicester, United Kingdom; 14 Icahn School of Medicine at Mount Sinai, Radiation Oncology, New York, USA; 15 University of Leicester, Genetics and Genome Biology, Leicester, United Kingdom; 16 Fundación Pública Galega , Medicina Xenómica, Santiago de Compostela, Spain; 17 Ghent University, Department of Human Structure and Repair, Ghent, Belgium; 18 University of Leicester, Leicester, United Kingdom., Genetics and Genome Biology, Leicester, United Kingdom; 19 German Cancer Research Center (DKFZ), Division of Cancer Epidemiology, Heidelberg, Germany; 20 University of Manchester, Translational Radiobiology Group, Division of Cancer Sciences,, Manchester, United Kingdom Purpose or Objective To use data from an international prospective cohort study of breast cancer patients (pts) to predict the risk of skin induration (SI) after radiotherapy (RT) using a machine learning algorithm that includes dosimetric/clinical/genetic factors. Materials and Methods Pts were treated after breast conserving surgery with conventional/moderate or ultra hypo-fractionated RT with or without a tumour bed boost based on clinical and pathological factors. Pts were enrolled in 7 countries in Europe/US; each centre followed local clinical practice, but the collection of data and genotyping was standardised and centralised. Our endpoint was late grade 1+ (G1+) SI 2 years after RT completion. Inclusion criteria were: no SI at baseline and availability of complete dosimetric and genetic data. For every pt, skin was defined as a 5-mm inner isotropic expansion from the outer body. To select a relevant portion of the skin DVH, we extracted the higher dose tail using different volume cutoffs (i.e. 25/50/100/150/200 cc volumes corresponding to 5x5-20x20cm2 areas). We corrected sub-DHVs for fractionation using two possible a/b values from the literature (1.8 Gy, Bentzen 1988 & Raza 2012; 3.6 Gy, Jones 2006 & Budach 2015). We calculated Equivalent Uniform Doses (EUDs) from corrected sub-DVHs, with n spanning from 1 to 0.05. We also considered the minimum dose of the selected DVH tail as an additional dose parameter (Dmin). Toxicity models were built using feed-forward neural networks (FNNs, 10 neurons and 1 hidden layer) following a wrapper method for feature selection. We used separate datasets for input: clinical/treatment/genetic features were constant, while the dosimetric factors (EUDs and Dmin) coming from sub-DVHs varied with volume cutoff and a/b (Fig 1). MO-0801 Machine learning based models of radiotherapy-induced skin induration for breast cancer patients