ESTRO 2025 - Abstract Book

S2778

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2025

for rectum, bowel and sigmoid; < 90Gy for bladder), which were strictly adhered to (2) . RO plans optimised all 10 CTVs (RO all) to achieve D95% > 22Gy and Dmax between 24 – 28Gy, prioritising central dose escalation for overlapping GTV volume. A practical approach was explored where a single, composite volume of all 10 CTVs was created (RO comp), optimised with the same planning aims. RO plans were normalised so that 9 out of 10 CTVs received D95% > 22Gy. Target coverage metrics (CTV D98%, D95% and the percentage of CTVs meeting D95%> 22Gy) and OARs doses were analysed. Statistical significance was assessed via one-way ANOVA, with PTV plans as controls. Results: Robust optimisation demonstrated significantly improved target coverage (Table 1). The proportion of CTVs meeting the coverage threshold was higher in RO plans (Figure 1) while cumulative dose to the dose-limiting OAR, the rectum, was significantly reduced; V60Gy, V65Gy and V70Gy.

Conclusion: The PTV approach fails to ensure adequate target coverage under DU, whereas RO improves target coverage and reduces OAR doses in high-dose regions. Incorporating DU from machine learning-based auto-contours could enable RO without IOV studies (3) . The RO composite approach offers a practical solution compatible with existing treatment planning systems.

Keywords: Robust optimisation, delineation uncertainty