ESTRO 2025 - Abstract Book

S2800

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

ESTRO 2025

Conclusion: RAD for esophageal cancer offers significant dosimetric advantages by reducing both lung and heart doses compared to standard VMAT, without compromise in PTV coverage. Additionally, RAD enhances planning efficiency, which may facilitate increased patient throughput, marking it as a promising option for high-quality and time effective esophageal cancer radiation therapy.

Keywords: RapidArc Dynamic, Dynamic Collimator, Esophagus

2006

Digital Poster From non-coplanar CyberKnife to coplanar Ethos in liver SBRT? Linda L Chen, Wilhelm den Toom, Abdul Sharfo, Linda Rossi, Ben Heijmen, Sebastiaan Breedveld, Yvette Seppenwoolde, Alejandra Méndez Romero Department of Radiotherapy, Erasmus MC Cancer Institute, Rotterdam, Netherlands Purpose/Objective: Liver SBRT is susceptible to variations in position and filling of the OARs between fractions. The Ethos system (Varian, a Siemens Healthineers Company) is designed to support a daily image-guided adaptive radiotherapy (ART) workflow to mitigate the effects of such variation. However, the Ethos is only suited for coplanar treatments, where our current protocol uses a non-coplanar IMRT treatment with the robotic CyberKnife (Accuray Corporation). As a first step to move towards ART, we have developed a workflow for automatic generation of VMAT plans for liver SBRT at an Ethos, and we have compared these plans with current manually generated clinical non-coplanar CyberKnife plans. Material/Methods: Following institutional review board approval, we included thirty-two patients with liver metastases, not eligible for surgery or ablation, and treated with 60 Gy in eight fractions between July 2018 and March 2024. Automated generation of VMAT plans was performed with our in-house developed multi-criterial optimizer [1]. First, eight patients were used to tune a wishlist for configuration of the automated multi-criterial plan optimization according to aims and trade-offs used in clinical planning. Then, autoVMAT plans were compared to clinical CyberKnife plans regarding PTV- and GTV coverages and doses delivered to the organs at risk (OARs), using clinical dose criteria. The Wilcoxon signed rank test was used for significance testing. Results: All autoVMAT plans adhered to clinical hard constraints. Figure 1 shows autoVMAT and CyberKnife plans for an example patient. AutoVMAT had on average a higher PTV coverage than CyberKnife (97.9% vs. 97.0%, p=0.002) and equal GTV coverage ( figure 2 ). AutoVMAT also had lower right kidney D 67% (0.0 vs. 0.8 Gy, p<0.001), lower duodenum D 0.01cc (3.9 vs. 10.3 Gy, p=0.019) and lower duodenum D 5cc (0.1 vs. 3.7 Gy, p<0.001) ( figure 2 ). On the other hand, autoVMAT had higher spinal cord D 0.01cc (13.4 vs. 7.5 Gy, p<0.001), higher esophagus D 0.01cc (18.4 vs. 12.9 Gy, p=0.002) and higher stomach D 5cc (14.1 vs. 12.3 Gy, p=0.012). For the other OARs plan parameters, differences were not statistically significant.