S1133
Clinical - Urology
ESTRO 2026
Material/Methods: Patients with localized prostate cancer (PCa), regardless of risk group or prostate volume, were treated with linac-based SBRT delivering 30 Gy in in three alternate-day fractions (EQD21.5:99Gy, EQD23: 78Gy), using 3-mm isotropic PTV margins. Real-time 4D-ultrasound prostate tracking (Clarity® Autoscan) enabled beam gating. All patients received prophylactic alpha-blockers. Maximum acute ( ≤ 90days) and late (>90days) SE (CTCAE v5.0), IPSS, EPIC-CP, IIEF 5, and PSA were collected at baseline and during follow-up. Statistical analyses included paired t- tests for patient reported outcomes (PROs) and univariate logistic regression for associations with
Digital Poster 252
Dose–volume signatures and toxicity after five- fraction prostateSBRT: PCA/ROC analysis with external benchmarking to PACE-B Amadeo Jesús Wals Zurita, Ana Illescas Vacas, María Rubio Jiménez, Paula Vicente Ruíz, Hector Miras del Río, Jonathan Saavedra Bejarano, Francisco Carrasco Peña, Antonio Ureña, Monica Ortíz Seidel Department of Radiation Oncology, Hospital Universitario Virgen Macarena, Seville, Spain Purpose/Objective: Five-fraction prostate SBRT is widely adopted, yet practical,data-driven DVH rules that generalize across centres remain limited. We analysedrectum/bladder and rectal/bladder wall dose–volume features using principal componentanalysis (PCA) and evaluated their ability to discriminate clinician-assessed GU/GI toxicityat 24 months with receiver operating characteristic (ROC) methods, benchmarking againstPACE-B two-year toxicity. Material/Methods: Retrospective single-centre cohort of localized prostate cancertreated with SBRT 36.25 Gy in 5 fractions. Valid DVHs were extracted for rectum/rectalwall, bladder/bladder wall, and targets (CTV/PTV). PCA summarised DVHs ( ≥ 90% varianceretained); Spearman correlations assessed associations between PTV and OAR Vxx. ROCanalyses (AUC) evaluated binary 24-month toxicity (yes/no). Longitudinal toxicity was alsosummarised. PACE-B served as an external benchmark for expected 2-year GU/GI toxicityrates. Results: SBRT 36.25 Gy/5 yielded low late toxicity with a subacute GU “flare” peaking at 6–12 months. At 24 months (n=58 evaluable), GU any-grade was 15.5% ( ≥ G2: 6.9%), and GI ≥ G1 was 3.4%; cumulative ever- events across follow-up (n=65) were GU 28.3% and GI14.2%. These patterns aligned with PACE-B.PCA indicated that low–mid dose regions carried most variance (e.g., rectum ~13 Gy;bladder ~6.5–34.5 Gy), while high-dose volumes were more constrained. Discriminativefeatures converged from two complementary analyses: (i) organ-based DVHs highlightedrectum V13% as the strongest 24-month GI discriminator (AUC ≈ 0.72), and (ii) wall-basedDVHs identified high-dose rectal wall exposure (e.g., PR_37.5) as the top GI signal(AUC ≈ 0.85 in our wall-analysis subset).For GU, bladder-wall high-dose metrics functioned mainly as rule-in/rule-out markers (e.g.,PV_37 with high NPV), and smaller bladder volumes were associated with GU events.Data-driven cut-points were stricter than guideline-oriented constraints and are bestframed as calibrated rule-out thresholds to complement planning
acute SE. Results:
Among 75 patients enrolled between November2023 and February2025, 46 had at least one late follow-up (median 9 months, range 5-15) and were included in this analysis. Median age was 79 years (56–85); 70% had intermediate-risk; 48% received ADT. Median CTV and PTV were 69cc (37–127) and 112cc (65–185), respectively. Acute GU SE were G1 in 22%, G2 in 35%, G3 in 2%, while late GU SE were G1 in 15%, and G2 in 9%. Acute GI SE occurred in 15% (G1) and 9% (G2), with late GI SE observed in 9% (G1).Acute GU SE was associated with larger CTV (p=0.017, 78vs60cc), longer time between simulation and treatment (p=0.023, 37vs29days), and higher baseline IPSS (p=0.039, 9vs5points). Late GU SE correlated with acute GU events (p=0.032), baseline IPSS (p=0.088, 10vs5) and acute QoL worsening (p=0.047, 17vs12). Acute GI SE correlated with larger rectal volumes (p=0.047, 86vs66cc). No predictor emerged for late GI.Changes in IPSS from baseline to the acute and late phases were not statistically significant. In contrast, QoL scores significantly worsened at both acute (p=0.005) and late follow-up (p<0.001).Median PSA dropped from 8.27ng/ml (2.5-49.5) at baseline to 0.66ng/ml (0.01-3.69) at last follow-up, with no biochemical failure events.Analysis of intra-fraction motion revealed displacements >2.5mm in 15% of sessions (3% transient, 9% non-transient, 3% mixed), predominantly in the posterior direction (p<0.001). These shifts triggered automated gating in 15% of fractions, enabling real-time beam interruption and re- alignment, thereby maintaining dosimetric accuracy and treatment safety. Conclusion: Our findings show that three-fraction SBRT with 4D- ultrasound monitoring is associated with low SE, optimal biochemical control and only modest QoL decline in patients with organ confined PCa. Longer follow-up is needed to confirm results Keywords: SBRT, prostate cancer, tracking
Made with FlippingBook - Share PDF online