ESTRO 2025 - Abstract Book

S357

Brachytherapy - Physics

ESTRO 2025

For a set of nine patients, previously treated with plans optimized using BrachyVision (Varian Medical Sys.) BRIGHT was used to generate new plans. The patients were selected to cover the typical range of CTV sizes and geometrical implant quality. The clinical protocol (see Table 1) contains target aims for the CTV_1cm and CTV_1.5cm. Dwell positions were activated automatically within a volume around the CTV_1.5cm. A ring structure was created outside of this activation volume in which V100% was minimized to enhance conformity. Dose inhomogeneity is strongly correlated with toxicity in breast. While not optimizable by BrachyVision, the Dose Homogeneity Index (DHI), , is evaluated after optimization. BRIGHT allows direct optimization of DHI as well as target aims, combined in a worst-case manner. A hard constraint on the ring V100% was used. This constraint is respected during optimization, but slightly higher values can result after re-evaluation on more dose calculation points (DCPs). We used a limit of 1cm3 to achieve similar values as observed in clinical plans. BRIGHT was configured to optimize on 120,000 DCPs for 60s. Final reported dosimetric indices were calculated using 500,000 DCPs. Results: Table 1 displays metrics for clinical plans and BRIGHT plans, selected following Figure 1. For patients 1 through 7, only small differences in coverage and DHI are observed, though BRIGHT improved both in six cases. For patients 3, 8, and 9, clinical plans had large non-protocol-satisfying deviations. For patients 3 and 9, BRIGHT improved ring V100% substantially at similar coverage and DHI. For patient 8, BRIGHT achieved coverage aims at similar DHI values. Conclusion: BRIGHT is capable of optimizing the trade-off between coverage and homogeneity when applied to breast brachytherapy, achieving results similar to or better than clinical practice while providing multiple plans to choose from. defined as