ESTRO 2020 Abstract book

S445 ESTRO 2020

Sacro Cuore, Istituto di Radiologia, Roma, Italy ; 5 San Raffaele Scientific Institute, Medical Physics, Milan, Italy

Purpose or Objective Early Regression Index (ERI TCP ) is a radiobiological parameter that showed promising results in predicting pathological complete response (pCR) on T2-weighted 1.5T MR images of patients affected by Locally Advanced Rectal Cancer (LARC). It models the early rectal cancer regression combining volumetric tumour information obtained during treatment simulation (V pre ) and at mid therapy (V mid ) (Formula in Fig.1). This study aims to validate ERI TCP in the context of low tesla MR-guided Radiotherapy (MRgRT), using MR images acquired with different magnetic field strength and imaging sequence. Furthermore, the optimal timing of MR imaging for pCR prediction was estimated, calculating the ERI index at different dose levels and quantifying its predictive ability in terms of AUC. Material and Methods 43 patients were enrolled from 3 institutions and subject to a MRgRT treatment delivered in 25-28 fractions with concomitant chemotherapy. Tumour regression grade (TRG) was defined according to Mandard and pCR was defined as TRG=1. For each patient, a true fast imaging with steady state precession sequence was acquired during the simulation and every day of the MRgRT treatment using a 0.35 T MR scanner. GTV contouring was shared between 2 radiation oncologists and blinded with respect to any information regarding the TRG. The physical doses were converted in Biologically Effective Dose (BED) to compensate the different radiotherapy schemes adopted in the centres, considering α/β equal to 10, recovery factor equal to 0.6Gy/day and proliferation delay time equal to 7 days. ERI TCP was calculated considering the fractions corresponding to a BED of 23Gy. Table 1 summarises the treatment details of the different centres and the fractions analysed to obtain the same BED levels. To investigate the optimal timing to predict pCR, different ERI values were estimated considering as V mid the volumes at different BED levels: 12, 36 and 45Gy.

Conclusion In conclusion, an online adaptive workflow with full replanning for hypofractionated radiotherapy in patients with rectal cancer on a 1.5T MR-linac is feasible, well tolerated and currently takes about 48 minutes per fraction.

Poster Highlights: Poster highlights 22 PH: Outcome modelling

PH-0715 External validation of ERITCP as response predictor in rectal cancer using MR-guided Radiotherapy D. Cusumano 1 , L. Boldrini 1 , P. Yadav 2 , Y. Gao 3 , G. Chiloiro 1 , A. Piras 4 , S. Broggi 5 , J. Lenkowicz 4 , L. Placidi 1 , H. Musunuru 2 , N. Dinapoli 1 , B. Barbaro 1 , L. Azario 1 , M.A. Gambacorta 1 , M. De Spirito 1 , M. Basetti 2 , Y. Yang 3 , C. Fiorino 5 , V. Valentini 1 1 Fondazione Policlinico Universitario A.Gemelli IRCCS, Dipartimento di Diagnostica per immagini- Radioterapia Oncologica ed Ematologia, Roma, Italy ; 2 School of Medicine and Public Health- University of Wisconsin- Madison, Department of Human Oncology, Madison, USA ; 3 University of California, Department of Radiological Sciences, Los Angeles, USA ; 4 Università Cattolica del

Results 7 patients showed pCR. ERI TCP showed the highest predictive performance, showing an AUC of 0.96 (95%CI:0.93-1). In particular, using the threshold of 13.1 previously suggested from the analysis of the training cohort, this parameter correctly classified 41/43 cases