ESTRO 2020 Abstract book

S447 ESTRO 2020

PH-0717 Locally Advanced Pancreatic Cancer: Robust Radiomic based model of outcome after radio- chemotherapy M. Mori 1 , P. Passoni 2 , E. Incerti 3 , S. Broggi 1 , G.M. Cattaneo 1 , M. Reni 4 , E. Spezi 5 , N. Slim 2 , E.G. Vanoli 3 , V. Bettinardi 3 , L. Gianolli 3 , M. Picchio 3 , N.G. Di Muzio 2 , C. Fiorino 1 1 IRCCS San Raffaele Scientific Institute, Medical Physics, Milan, Italy ; 2 IRCCS San Raffaele Scientific Institute, Radiotherapy, Milan, Italy ; 3 IRCCS San Raffaele Scientific Institute, Nuclear Medicine, Milan, Italy ; 4 IRCCS San Raffaele Scientific Institute, Oncology, Milan, Italy ; 5 Cardiff University, School of Engineering, Cardiff, United Kingdom Purpose or Objective PET Radiomics is a promising tool to predict outcome after Radiotherapy (RT) although no reliable results exist for Locally Advanced Pancreatic Cancer (LAPC). The aim of the study was to develop robust radiomic based predictive models for LAPC treated with exclusive chemo-RT. Poorly repeatable/reproducible PET Radiomic Features (RF) were previously identified and withdrawn. Material and Methods The study involved a single-center cohort of 120 LAPC patients (pts) treated with chemo-RT within a study investigating moderate hypo-fractionation (44.25 Gy in 15 fr, with/without boosting PET positive volumes to 48-54 Gy). The IBSI (International Biomarker Standardization Initiative) compliant software SPAARC was used to extract RF. ROC and COX regressions were performed to detect the predictive power of the most robust RF for Overall, Local and Distant-Relapse Free Survival (OFS, LRFS and DRFS). COX regression multi-variable models including RFs with/without including available clinical parameters were generated by first eliminating redundant RFs, then by backward selection of best predictive parameters and finally by bootstrap-based validation. Predictive indices based on the COX models for OS, LRFS and DRFS were derived as the cut-off values best discriminating pts with the event (death or relapse) at the time corresponding to the 1 st quartile (6, 5, and 2,5 months from RT start for OS, LRFS and DRFS respectively).