ESTRO 2025 - Abstract Book

S4312

RTT - Treatment planning, OAR and target definitions

ESTRO 2025

Material/Methods: Materials/methods: A total of 90 simulation-free treatments were simulated (based on 27 patients with 30 target sites). Diagnostic CTs from patients previously treated for MSCC were used for initial treatment planning. The first CBCT was then used for oART simulation of this plan. Three simulation-free workflows were defined and assessed: rigid contour propagation (RCP), deformable contour propagation (DCP), and an unsupervised workflow (deformable propagation without manual corrections). The simulations were performed in the Ethos emulator, a virtual copy of the clinical Ethos treatment system. Time stamps, contours, and dose plans were used to evaluate the feasibility of each workflow. Results: Results: Median time consumption (from CBCT acceptance to treatment start) was 7.49 min., 6.57 min., and 3.27 min. for the RCP, DCP, and unsupervised UCP workflows, respectively (Figure 1). The CTV structures from the three adaptive workflows showed a high dice similarity coefficient compared to the CTV structures from the diagnostic CTs (Figure 2). All adaptive plans from the three workflows met all clinical goals for dose coverage. Significant differences were observed when comparing the dose from the adaptive plans to the dose from the scheduled plans (p<0.05).

Figure 1.

Figure 2.

Conclusion: Conclusion: The unsupervised workflow demonstrated superior speed compared to the RCP and DCP workflows. While the workflow exhibited a high degree of similarity in contouring, significant differences were still observed in a few cases, such as variations in target placement. Therefore, further investigation is necessary to fully assess its potential. The DCP workflow was found to be the most optimal and safe for clinical implementation, as demonstrated by the successful treatment of the first patient with MSCC using this approach.