S2764
RTT - RTT contouring, target definition, and treatment planning
ESTRO 2026
(OAR) sparing of Raystation’s AI module (rayAutoBreast) against manual field-in-field (FiF) 3D- CRT planning for right-breast cancer patients after breast-conserving therapy, stratified by breast size (small vs. large). Material/Methods: A cohort of 30 right-breast cancer patients, with 15 patients classified as small-breast (field separation <22 cm) and 15 as large-breast (field separation ≥ 22 cm). Treatment planning was performed using manual Field-in-Field (FiF) plans with Eclipse v15.6 TPS and AI- generated plans utilizing Raystation’s rayAutoBreast (v12B). The prescribed dose was 40.05 Gy in 15 fractions, with organs at risk (OARs) including the hearts, ipsilateral lungs, contralateral lungs, and contralateral breasts. Dosimetric evaluation focused on target coverage metrics such as Conformity Index (CI), Homogeneity Index (HI), D98% and D2% of the target volumes, as well as V105%. OAR sparing was assessed through mean heart dose (MHD), heart V20Gy, lung V4Gy, V8Gy, V16Gy, V16.683Gy, mean lung dose (MLD), and contralateral breast maximum dose. Wilcoxon signed-rank tests were employed to test if there was any significant dosimetric difference between AI and manual plans. Dosimetric improvement was also compared between large and small breast groups using Mann-Whitney U tests. Results: Results demonstrated that AI plans significantly improved target conformity with a reduction in CI for both large (1.57 vs. 1.85, p=0.003) and small breast groups (1.84 vs. 2.17, p=0.021). Although V105% increased with AI in small breast group (85.5 cm ³ vs. 18.7 cm ³ , p=0.004), it remained clinically acceptable, and homogeneity indices were comparable between methods. AI also showed superiority in OAR sparing, notably lowering mean heart dose in the large-breast group (0.47 Gy vs. 0.80 Gy, p=0.035), reducing V4Gy for the ipsilateral lung in both groups, and decreasing V16.683Gy and MLD in lung tissues. Comparing with small-breast patients, the benefits of AI were more pronounced in large-breast patients, with significant improvements in CI, MHD, and ipsilateral lung V4Gy ( Δ CI: 0.29 vs. 0.23, p=0.043; Δ MHD: 0.33 Gy vs. 0.14 Gy, p=0.039; Δ V4Gy: 7.52% vs. 4.9%, p=0.043), highlighting the potential for AI-driven planning to optimize dosimetric outcomes, especially in larger breast volumes. Figures below are examples to illustrate the dosimetric comparison between two methods for patients with different breast sizes.
Conclusion: Raystation’s rayAutoBreast enhances dosimetric performance for right-breast radiotherapy, particularly in large-breast patients where anatomical complexity challenges manual planning. AI achieves superior PTV conformity, reduced heart and lung doses. This demonstrates AI’s potential to standardize high-quality planning and mitigate OAR toxicity risk in breast cancer radiotherapy. References: Duzenli, C., Koulis, T., Menna, T., Carpentier, E., Arora, T., Coope, R., Gill, B., Lim, P., Aquino-Parsons, C., Nichol, A., Singer, J., Ingledew, P. A., Grahame, S., & Chan, E. K. (2021). Reduction in Doses to Organs at Risk and Normal Tissue During Breast Radiation Therapy With a Carbon-Fiber Adjustable Reusable Accessory. Practical radiation oncology, 11(6), 470–479. https://doi.org/10.1016/j.prro.2021.06.012 Keywords: Breast cancer, AI planning, dosimetric comparison Comparison of pencil beam scanning proton therapy techniques for reduction of mandibular dose in locally advanced oropharyngeal cancer Xinying Lin 1,2 , Michelle Leech 2,3 1 Division of Radiation Oncology, National Cancer Centre Singapore, Singapore, Singapore. 2 Applied Radiation Therapy Trinity, Discipline of Radiation Therapy, Trinity College Dublin, Dublin, Ireland. 3 Applied Radiation Therapy Trinity, Discipline of Radiation Therapy, Trinity St. James’s Cancer Institute, Dublin, Ireland Purpose/Objective: Osteoradionecrosis (ORN) is a serious late toxicity from radiotherapy of head and neck cancer and occurs most commonly in the mandibular bone in patients treated with radiotherapy for oropharyngeal cancer. Proton therapy offers potential for improved normal tissue sparing, but treatment techniques of oropharyngeal cancer are currently varied across studies. The aim of this project is to compare the mandible dose from two pencil beam scanning (PBS) techniques when planning locally advanced oropharyngeal cancer patients treated with proton therapy. Digital Poster 510
Made with FlippingBook - Share PDF online