ESTRO 2023 - Abstract Book

S1607

Digital Posters

ESTRO 2023

Conclusion Inter-fractional variation in organ filling (esp. bladder) is inevitable despite fixed pre-treatment protocol in definitive LACC. Variation in bladder filling is patient-specific, multi-factorial and has a systemic component. Despite the logistical challenges and inherent heterogeneity in the selection of ‘plan of the day’; adaptive IG-IMRT based on incremental CTV to-PTV margins is a feasible and promising strategy to minimize geometric uncertainties in radical RT of cervical cancer.

PO-1865 Benefit of daily online cbct-based adaptive radiotherapy in prostate cancer: a preliminary analysis.

A. D'Aviero 1 , M. Aquilano 1 , A. Boschetti 1 , F. Catucci 1 , P. Gaias 1 , F. Gruosso 1,2 , M. Iezzi 1 , M. Marras 1 , S. Menna 3,4 , D. Piccari 1,2 , E. Pilloni 3 , D. Piro 1,4 , F. Preziosi 1 , F. Quaranta 3 , A. Re 1 , V. Verusio 5 , C. Votta 1,2 , V. Valentini 2,5 , D. Cusumano 3,2 , G.C. Mattiucci 1,5 1 Mater Olbia Hospital, Radiation Oncology, Olbia, Italy; 2 Fondazione Policlinico Universitario Agostino Gemelli IRCCS, UOC Radioterapia Oncologica, Roma, Italy; 3 Mater Olbia Hospital, Medical Physics, Olbia, Italy; 4 Fondazione Policlinico Universitario Agostino Gemelli IRCCS, UOC Radioterapia Oncologica, Olbia, Italy; 5 Università Cattolica del Sacro Cuore, Dipartimento di Diagnostica per Immagini, Radioterapia Oncologica ed Ematologia, Roma, Italy Purpose or Objective Aim of this preliminary analysis was to evaluate the potentialities in terms of dosimetric benefit of online daily adaptation in prostate cancer by using a modern CBCT-based linac able to provide online adaptive radiotherapy (online ART) using Artificial Intelligence (AI) system Materials and Methods A total of 11 prostate cancer patients treated with online ART were enrolled for this study. The patients underwent radiotherapy treatment on an AI-based Linac (Ethos, Varian, US) with a total dose of 67.5 Gy administered in 25 fractions. CTV was defined as prostate and seminal vesicles according to clinical stage. An isotropic margin of 5 mm was added to CTV to obtain PTV except for caudal margin with 0.8 mm expansion. A daily CBCT was acquired for each treatment session, and a daily online adaptive workflow was performed with the collaboration among a radiation oncologist (RO), a medical physicist and a radiation therapist (RTT). The AI-system automatically segmented organs at risks (OARs) consisting in rectum, bladder and bowel with an online check by the RO, target volumes (CTV and PTV) were manually delineated. Treatment plan was automatically reoptimized by the system and two treatment plans were obtained: a predicted plan, consisting in the calculation of the original fluence on the daily anatomy, and an adapted plan, result of a new optimisation. Dose Volume Histogram for both adapted and predicted plan were obtained daily including clinical target coverage values for PTV (V95% and V105%), also reporting indicators for organs at risk such as bladder (V65Gy), bowel (V45Gy) and rectum (V50Gy).

The difference among predicted and adapted indicators were investigated using the Wilcoxon Mann Whitney test for paired sample. Significance was considered in case of p-value less than 0.05.

Results A total of 193 fractions were analysed. For predicted plans mean V95% of PTV was 86.4%, while following online adaptation a mean value of 98.4±1.1% was obtained with a statistically significant (p<0.01) difference between the two modalities.

As regards OARs, no significant difference was observed in terms of dose sparing (probably due to the patient preparation