ESTRO 2021 Abstract Book

S1118

ESTRO 2021

risk of recurrence in univariate analysis (HR 1.9, 0.63-5.8, p=0.26). The 5-year DFS and MFS were 100% for PI- RADS 3 patients, 93% and100% for PI-RADS 4, and 82% and 91% for PI-RADS 5, respectively. Notably, all of the distant failures occurred in PI-RADS 5 patients. Conclusion We found a trend towards a higher rate of treatment failure in patients with high PI-RADS scores. All of the distant failures and majority of biochemical failures occurred in patients with a PI-RADS score of 5. Further studies analysing larger cohorts are needed. PO-1363 Prediction of urinary toxicity after prostate cancer RT: long term evaluation T. Rancati 1 , A. Cicchetti 1 , E. Gioscio 1 , B. Avuzzi 2 , B. Noris Chiorda 2 , R. Valdagni 3 , E. Garibaldi 4 , D. Cante 5 , V. Vavassori 6 , G. Girelli 7 , C. Degli Esposti 8 , C. Iotti 9 , C. Fiorino 10 , C. Cozzarini 11 1 Fondazione IRCCS Istituto Nazionale dei Tumori, Prostate Cancer Program, Milan, Italy; 2 Fondazione IRCCS Istituto Nazionale dei Tumori, Radiation Oncology 1, Milan, Italy; 3 Fondazione IRCCS Istituto Nazionale dei Tumori and Università degli Studi di Milano, Prostate Cancer Program and Radiation Oncology 1, Milan, Italy; 4 Istituto di Candiolo-Fondazione del Piemonte per l'Oncologia IRCCS, Radiation Oncology, Candiolo, Italy; 5 Ospedale di Ivrea, Radiation Oncology, Ivrea, Italy; 6 Cliniche Gavazzeni-Humanitas, Radiotherapy, Radiation Oncology, Bergamo, Italy; 7 Ospedale degli Infermi, Radiation Oncology, Biella, Italy; 8 Ospedale Bellaria, Radiation Oncology, Bologna, Italy; 9 ASMN Hospital IRCCS, Oncology and Advanced Technology, Reggio Emilia, Italy; 10 IRCCS Istituto Scientifico Ospedale San Raffaele, Medical Physics, Milan, Italy; 11 IRCCS Istituto Scientifico Ospedale San Raffaele, Radiation Oncology, Milan, Italy Purpose or Objective To assess factors affecting the incidence of (patient-reported) urinary toxicity (tox) at 5 years after radical radiotherapy (RT) for prostate cancer in a large group of patients (pts) enrolled in a prospective, multicentric trial Materials and Methods Pts were treated at different prescribed doses with conventional (74-80Gy @1.8-2Gy/fr) or moderately hypo- fractionated RT (65-75.2Gy @ 2.2-2.7Gy/fr) in 5fractions/week. All doses were corrected to 2Gy/fr, using the linear-quadratic model by applying α/β derived from maximum likelihood estimate (MLE). Bladder dose- volume histogram were reduced to equivalent uniform dose (EUD), with volume parameter n derived from MLE. Tox was evaluated every 6 months till 60 months. 4 endpoints were considered: (a) urinary incontinence as evaluated through the International Consultation on Incontinence Modular Questionnaire short form (ICIQ) and defined as the occurrence of an ICIQ value>5 (as calculated in the first 2 questions: frequency of urinary incontinence & amount of leakage) at least once between 6 and 60 months. Pts with baseline ICIQ>0 in the first 2 questions were excluded from the analysis. (b) urinary obstructive symptoms as evaluated with the International Prostate Symptom Score (IPSS) questionnaire and defined as the occurrence of an increase of IPSS of at least 10 points with respect to baseline at least once between 6 and 60 months. (c) grade 2+ haematuria as scored by the clinician. (d) obstruction requiring intervention (e) a comprehensive moderate/severe toxicity defined as at least one among a/b/c/d. Logistic was used to determine association between urinary tox probability and the best dosimetric descriptor (determined MLE) and clinical risk factors. Internal validation was performed by bootstrapping. Nomograms were derived from logistic Results 312 pts were available. Rate for incontinence was 14%, 16.5% for obstructive symptoms, 3% haematuria, 2% obstruction and 26.6% for any moderate/severe tox. With the only exception of haematuria, for all the other tox endpoints MLE pointed toward the 2Gy-equivalent prescribed dose, with alpha/beta=1.5Gy, indicating a high sensitivity of urinary tox to fractionation and a possible relevant role of the urethra for these endpoints. Odds Ratios (OR) were 1.06, 1.09, and 1.06 for 1Gy increase for incontinence, obstructive symptoms and any tox, respectively. For haematuria the dose to 2cc of bladder (α/β=5Gy) was selected by MLE (OR=1.23). Some clinical features were significantly associated to the different endpoints as reported in nomograms in fig.1. It was not possible to model obstruction separately due to the low number of events