ESTRO 2026 - Abstract Book PART I

S1216

Clinical - Urology

ESTRO 2026

Milan, Italy. 10 Medical Physics Unit, Cliniche Humanitas-Gavazzeni, Bergamo, Italy. 11 S.C.

Results: 388 patients were available. 23/388 presented urinary_incontinence pre-radiotherapy:365 evaluable patients, 81/365 (22.2%) experiencing urinary_incontinence.MLE yielded bladder seriality parameter n=0.01 and α / β =0.50Gy, both at the lower bound of optimisation range. Estimates of n and α / β were consistent across all configurations, regardless of the combination of risk factors included in the MLE- fit.The fitted dose–response parameters corresponded to EUD_50=114.22Gy and m=0.18. Three clinical factors were included: diabetes (DMF=0.91), previous abdominal surgery (DMF=0.89), and TURP (DMF=0.89). Figure_1 reports model’s coefficients, NTCP curves, calibration plot. Figure_2 presents an association tree with hierarchical comorbidity co-occurrence patterns and observed/predicted urinary_incontinence rates in patients’ subgroups. Conclusion: Late urinary_incontinence followed a highly serial-like and steep dose–response pattern, with pronounced sensitivity to fractionation. Results are comparable for grade2+ urinary_incontinence models from CHHiP (EUD_50=109Gy, m=0.24, n=0.02, α / β =1.5Gy with confidence_interval 0.1-6.2Gy, Brand2022) and Requite (EUD_50=138.5Gy, m=0.33, n=0.07, α / β =0.07Gy, Rancati2022).Diabetes, previous abdominal surgery, and TURP highly modulate the risk, suggesting the need of personalized dose optimization goals.

Radioterapia Oncologica, PO Umberto Parini, Aosta, Italy. 12 Radiation Oncology Unit, ASL TO4 Ospedale di Ivrea, Ivrea, Italy. 13 Radiation Oncology Unit, Ospedale Maggiore della Carità di Novara, Novara, Italy. 14 Radiation Oncology Unit, Cliniche Humanitas- Gavazzeni, Bergamo, Italy. 15 Radiation Oncology Unit, Ospedale di Biella, Biella, Italy. 16 Radiation Oncology Unit, Ospedale Bellaria, Bologna, Italy. 17 Radiation Oncology Unit, AUSL- IRCCS di Reggio Emilia, Reggio Emilia, Italy. 18 Medical Physics Unit, IRCCS Istituto Scientifico San Raffaele, Milan, Italy. 19 Radiation Oncology Unit, IRCCS Istituto Scientifico San Raffaele, Milan, Italy Purpose/Objective: Model the dose–response relationship for late urinary_incontinence after radical prostate cancer radiotherapy by estimating the Equivalent Uniform Dose (EUD)-based bladder seriality and fractionation effects, while also incorporating patient-specific factors. Material/Methods: We considered patients enrolled in a prospective multicentre observational study, treated with varying doses using conventional (74-80Gy@1.8-2Gy/fraction) or moderate hypo-fractionation (65-75.2Gy@2.2- 2.7Gy/fraction).Urinary_incontinence was assessed using the International Consultation Incontinence Questionnaire (ICIQ-SF). Following LAPERS (Late, Persistent, Substantial, Treatment-Related Symptoms, Kirchheiner2020), patients were considered to have clinically relevant urinary_incontinence over 5-year follow-up if reported median urinary_incontinence level ≥ 2 in ICIQ-SF question_3, indicating experiencing urine leakage at least two/three times/week, and with worsening compared to pre-radiotherapy. Patients with median levels ≥ 2 but not worsening were excluded.To estimate the dose-response relationship, we used maximum likelihood estimation (MLE), with logistic function. EUD (dosimetric predictor) was computed from bladder dose–surface histograms and corrected for fractionation using linear–quadratic model. We fitted α / β , bladder volume-seriality effect (n), and dose-modifying effects for clinical factors within MLE. Model robustness was evaluated by 1000 repeated MLEs on jackknife samples (90% of original dataset), separately resampled for events/non-events. Parameter optimization was performed through a two-stage strategy, combining broad grid exploration with local refinement to ensure global convergence. Clinical factors were retained based on the likelihood- ratio test, and their effect included as dose-modifying factors (DMFs) acting on EUD_50 (bladder_EUD corresponding to 50% risk of urinary_incontinence, Defraene2012).

Made with FlippingBook - Share PDF online